betaFeedback

WorksMathematics

Orbit stratifications and smooth topology of positively curved six-manifolds with rank-one nonabelian symmetry

Abstract

We study closed simply connected six-manifolds with positive sectional curvature and an effective isometric action of SO(3)SO(3) or SU(2)SU(2). Starting from Liu's corrected orbit-space analysis, we refine the stratification and recover smooth topology in several previously unresolved cases. For a boundaryless SO(3)SO(3) quotient, singular isotropy is connected, the weighted nonprincipal graph belongs to nine necessary families, and every action with only cyclic exceptional isotropy is classified equivariantly. The remaining branching families admit explicit rational intersection and Pontryagin formulas. For a three-ball SO(3)SO(3) quotient, the uniform-boundary configurations give sphere bundles over S4S^4 and a completely distinguished odd-integer family of non-spin six-manifolds. A doubled-quotient argument excludes the configuration with two boundary fixed points and an interior singular orbit, while the configuration with one fixed point and an exceptional interval has underlying manifold CP3\mathbb{CP}^3. For a boundaryless SU(2)SU(2) quotient, an adjoint-weight matching argument reduces the exceptional three-orbit case to a unique action on the Eschenburg flag and identifies every two-orbit candidate with a weighted action on S2×S4S^2\times S^4. These classifications imply rational ellipticity throughout the stated symmetry class. We also establish restricted curvature obstructions for the two diagonal product actions, without asserting nonexistence for arbitrary invariant metrics. Abstract graph restrictions, smooth realizability, and positive-curvature realizability are kept distinct. Some of the results in this paper are discovered by ChatGPT-6-astra and verified by the authors.

Introduction

The relation between positive sectional curvature and topology remains one of the central questions in global Riemannian geometry. Compact rank-one symmetric spaces provide the basic examples. The homogeneous constructions of Wallach and the biquotient constructions of Eschenburg showed that this class is not exhausted by symmetric spaces [51, 9]. In dimension six the homogeneous flag manifold

FW=SU(3)/T2F_{\mathrm{W}} = \mathrm{SU}(3)/T^2

and its inhomogeneous Eschenburg counterpart FEF_{\mathrm{E}} are particularly important: they have the same Betti numbers, but their integral intersection data and Pontryagin classes distinguish them. Thus even a sharp Euler-characteristic theorem does not settle the smooth classification problem. Surveys of the examples and construction techniques can be found in [55, 54].

The symmetry program asks how much topology is forced when a positively curved metric has a prescribed isometry group. Grove’s formulation [17] encompasses both large torus rank and small cohomogeneity. The maximal-rank theorem of Grove–Searle [20], the almost maximal-rank result of Fang–Rong [12], Wilking’s connectedness and symmetry theorems [52, 53], the four-dimensional results of Hsiang–Kleiner and Grove–Wilking [31, 24], the five-dimensional result of Rong [46] and the recent ten-dimensional result of Krishnan–Wiemeler [38] demonstrate the strength of this approach. Nonabelian symmetry supplies different information from torus rank. A compact connected nonabelian Lie group contains a subgroup locally isomorphic to SU(2)\mathrm{SU}(2); its effective image is SU(2)\mathrm{SU}(2) or SO(3)\mathrm{SO}(3). These groups have rank one, so their presence alone is not a large-torus-rank hypothesis. In dimension six, however, a generically free action of either group has cohomogeneity three. The geometry of a three-dimensional quotient and the topology of six-manifolds can therefore be used together.

Related five-dimensional work includes Simas’s study of nonnegatively curved five-manifolds with nonabelian symmetry [47]. Liu subsequently established an Euler-characteristic theorem, determined the nontrivial-principal-isotropy actions, classified several fixed-point and exceptional-free cases, and analyzed the remaining orbit strata in dimension six [39, 40]. His corrected manuscript is our starting point. In particular, the polyhedral exceptional isotropy exhibited there must be retained: the irreducible action on the round harmonic-cubic six-sphere has tetrahedral isotropy. An argument that excludes every noncyclic exceptional group would therefore eliminate a genuine positively curved example.

Several recent developments clarify the setting of the present work. Kennard–Wiemeler–Wilking’s splitting methods and their connection between torus representations and regular matroids provide new curvature obstructions under higher-rank torus symmetry [34, 35]. These results concern a different symmetry regime from the rank-one actions considered here. More directly, Khalili Samani–Radeschi proved rational ellipticity for positively curved cohomogeneity-three manifolds with boundaryless orbit space and established a global catalogue of quotient graphs, together with branched-cover and unknotting tools [37]. Their results turn local slice restrictions into a finite list of abstract graph shapes. The double disk-bundle theorem of DeVito–Galaz-García–Kerín supplies a complementary route to rational ellipticity for the ball-quotient configurations [8]. We use these theorems as inputs.

Recent construction preprints also counsel care in formulating product obstructions. In August 2026 Brendle–Hung reported a positive-sectional-curvature construction on S2×S2S^2 \times S^2 using a higher-order perturbation of a Cheeger deformation [6], and Guo–Fang–Lu reported an analogous construction on S2×S3S^2 \times S^3 [27]. In September 2026 Guo–Fang–Lu reported a construction on S3×S3S^3 \times S^3 [28], and He–Liu–Yau reported constructions on all smooth homotopy seven-spheres [29]. These are cited here as recent preprints, not as ingredients in any proof below. They continue a construction direction exemplified earlier by Dearricott and Grove–Verdiani–Ziller [7, 23]. Neither report decides positive curvature on S2×S4S^2 \times S^4 with either of the specific rank-one actions studied here. In particular, the Euler-characteristic form of the Hopf conjecture, the historical product question, and a prescribed-action metric-existence problem are three different assertions. The last must be addressed by an obstruction to all invariant metrics, not merely to product or connection metrics.

Standing hypotheses and notation

Unless a statement explicitly concerns smooth realizability without curvature, we assume

M6 is a closed, smooth, simply connected manifold,M^{6}\text{ is a closed, smooth, simply connected manifold,}
g is a Riemannian metric with sec⁡g>0,g\text{ is a Riemannian metric with }\sec_g>0,
G=SO⁡(3) or SU⁡(2) acts effectively and isometrically on (M,g).(1)G=\operatorname{SO}(3)\text{ or }\operatorname{SU}(2)\text{ acts effectively and isometrically on }(M,g). \tag*{(1)}

Write X=M/GX=M/G and π:M→X\pi:M\to X. A singular orbit has dimension smaller than a principal orbit; an exceptional orbit has principal dimension but larger isotropy. Thus a finite-isotropy branching vertex is exceptional, not singular in this terminology. Let Σ⊂X\Sigma\subset X be the image of all nonprincipal orbits. An edge labelled mm means cyclic isotropy CmC_m of order m≥2m\ge2. A label 1 is omitted, since it describes principal points. We denote by Dm⊂SO⁡(3)D_m\subset\operatorname{SO}(3) the rotational dihedral group of order 2m2m.

For G=SO⁡(3)G=\operatorname{SO}(3) put T=SO⁡(2)T=\operatorname{SO}(2); for G=SU⁡(2)G=\operatorname{SU}(2) put

T={diag⁡(z,z−1):z∈S1}.T=\{\operatorname{diag}(z,z^{-1}):z\in S^1\}.

The real oriented weight-jj plane is denoted by CjC_j. Unless signs are needed for a normal Euler class, slice weights are positive unordered pairs. Cohomology has integral coefficients unless otherwise indicated. A statement about a diffeomorphism is an ordinary smooth classification unless equivariance is explicitly specified. All references to Liu’s theorem, remark, or figure numbers mean the corrected version [40], not the earlier published text [39].

The classification results

The following statements summarize the main conclusions. Their proofs occupy § 3–§ 5; the models and all parameters are defined there.

Theorem 1.1 (SO(3) and a spherical quotient). *Assume (1.1), G=SO⁡(3)G=\operatorname{SO}(3), and X≅S3X\cong S^3. Every singular isotropy group is TT, and the number rr of singular orbits satisfies 1≤r≤31\le r\le3 and χ(M)=2r\chi(M)=2r. The abstract weighted graph Σ\Sigma is one of the nine families in Table 1 and Table 2.

If every exceptional isotropy group is cyclic, then r=2r=2 or 33. For r=2r=2 the action is equivariantly diffeomorphic to

Np,qG=G×TS(R⊕Cp⊕Cq),p,q≥1,gcd⁡(p,q)=1,N^G_{p,q}=G\times_T S(\mathbb{R}\oplus C_p\oplus C_q),\qquad p,q\ge1,\qquad\gcd(p,q)=1,

and M≅S2×S4M\cong S^2\times S^4. For r=3r=3 it is equivariantly diffeomorphic to

Pa,b=G×TCP2=P(1⊕L2a⊕L2(a+b))⟶S2,a,b≥1,gcd⁡(a,b)=1,P_{a,b}=G\times_T\mathbb{C}P^2=\mathbb{P}(1\oplus L^{2a}\oplus L^{2(a+b)})\longrightarrow S^2,\qquad a,b\ge1,\qquad\gcd(a,b)=1,

where c1(L)c_1(L) generates H2(S2)H^2(S^2). The underlying manifold is the product S2×CP2S^2\times\mathbb{C}P^2 precisely when a≡b(mod3)a\equiv b\pmod3; otherwise it is the nontrivial unoriented diffeomorphism type of a linear CP2\mathbb{C}P^2-bundle over S2S^2.

The remaining graphs are a dihedral YY, five types of lollipop, and one D3D_3–A4A_4 theta graph. Their rational intersection forms and Pontryagin covectors are given in § 3.5. The nine-family list is necessary; it is not a positive-curvature realization theorem or a classification of all spatial embeddings.

TypeShapeWeights and isotropyrrχ\chi
ITwo pointsBoth slices (1,1)(1,1).24
IIIntervalEdge m≥2m \ge2; endpoint slices (1,m)(1,m).24
IIIBigonEdges m,n≥2m,n \ge2, gcd⁡(m,n)=1\gcd(m,n)=1; both slices (m,n)(m,n).24
IVPathEdges m,m+1m,m+1, m≥2m \ge2.36
VTriangleEdges a,b,a+ba,b,a+b, a,b≥2a,b \ge2, gcd⁡(a,b)=1\gcd(a,b)=1.36
VIPoint and intervalIsolated slice (1,1)(1,1) and one order-two interval.36
VIIYYFinite vertex DmD_m; arms 2,2,m2,2,m, m≥2m \ge2.36
VIIILollipopCycle orders 2,k2,k, stem ℓ\ell; finite vertex as in Table 2.24
IXSubdivided thetaFinite vertices D3,A4D_3,A_4; direct edges 2,32,3; third path D3→2T→3A4D_3 \xrightarrow{2} T \xrightarrow{3} A_4.12

Table 1. Necessary abstract graphs for SO(3) with spherical quotient. A leaf on an order-mm edge has slice (1,m)(1,m); a two-valent circle vertex has its two incident orders as slice weights.

Finite vertexCycleStemRequired finite-slice character
DkD_k, k≥3k \ge3 odd2,k2,k2Order-kk rotation weight ±2\pm2 mod kk.
A4A_42,32,33Usual tetrahedral slice.
S4S_42,32,34Usual octahedral slice.
A5A_52,32,35Order-five weight ±1\pm1.
A5A_52,52,53Order-five weight ±2\pm2.

Table 2. The five lollipop possibilities in VIII. The circle vertex on the cycle has slice (2,k)(2,k) and the leaf has slice (1,ℓ)(1,\ell). In VII the order-mm transverse slice weight at DmD_m is ±1\pm1.

Theorem 1.2 (SO(3) and a ball quotient). Assume (1.1), G=SO⁡(3)G=\operatorname{SO}(3), and X≅B3X\cong B^3.

  1. In Liu’s Figure 2(b), the underlying manifold is

En=S(ξn)⟶S4,p1(ξn)=4nu,n∈Z,E_n=S(\xi_n)\longrightarrow S^4,\qquad p_1(\xi_n)=4nu,\qquad n\in\mathbb{Z},

where ξn\xi_n is an oriented three-plane bundle and uu generates H4(S4)H^4(S^4). The ordinary diffeomorphism type is determined by ∣n∣|n|; E0=S2×S4E_0=S^2\times S^4 and E1=CP3E_1=\mathbb{C}P^3.

  1. In Figure 2(c), the smooth action is obtained by gluing a weight-(1, 1) singular-orbit tube to a two-disk bundle over S2×S2S^2\times S^2 whose Euler class is 2a+kb2a+kb. The gluing exists exactly

for odd kk and has no further equivariant attaching parameter. The resulting manifolds MkM_k satisfy

Mk≅Mℓ⟺∣k∣=∣ℓ∣.M_k \cong M_\ell\Longleftrightarrow|k| = |\ell|.

They are non-spin, have torsion-free homology, and their full cubic and Pontryagin data are given in (4.5)–(4.7).

(iii) Figure 2(e) cannot occur. More generally, if ff is the number of isolated boundary fixed points with tangent representation R3⊕R3\mathbb{R}^3 \oplus\mathbb{R}^3, and rr is the number of interior singular points, then f+2r≤3f + 2r \le3.

(iv) In Figure 2(f), MM is diffeomorphic to CP3\mathbb{C}\mathrm{P}^3. The known projective SO(3)\mathrm{SO}(3) action realizes this configuration with positive curvature.

Figures 2(a) and 2(d) are the cases treated by Liu; in (d) the underlying manifold is diffeomorphic, not merely homeomorphic, to the standard S6S^6. The smooth families in (i)–(ii) do not assert positive-curvature realizability for every parameter.

Theorem 1.3 (SU(2) and a spherical quotient). Assume (1.1), G=SU(2)G = \mathrm{SU}(2), and X≅S3X \cong S^3. Then the action is one of the following equivariant diffeomorphism types:

(i) the linear action on S(C2⊕su(2))≅S6S(\mathbb{C}^2 \oplus\mathfrak{su}(2)) \cong S^6;

(ii) N1,1GN^G_{1,1} or N2j−1,2j+1GN^G_{2j-1,2j+1}, j≥1j \ge1, on S2×S4S^2 \times S^4;

(iii) the standard block action on FWF_{\mathrm{W}};

(iv) the commuting right block action on FEF_{\mathrm{E}}.

The only genuine exceptional stratum in the three-singular-orbit case is one C3C_3 interval, and its presence characterizes (iv). Consequently the middle panel of Liu’s Figure 4 has labels (1,1)(1,1) or (1,3)(1,3), and the right panel has (1,1,1)(1,1,1) or (1,1,3)(1,1,3), up to permutation and deletion of labels 1. The surviving three-orbit patterns have positive-curvature realizations. For the two-orbit patterns, the displayed weights are necessary conditions; invariant strict positivity is not established here.

The new geometric input in Theorem 1.3 is a matching condition between an exceptional slice and the adjoint representation: if a CkC_k edge has transverse character ss, positive curvature forces s≡±2(modk)s \equiv\pm2 \pmod{k}. The proof uses Killing Jacobi fields and the fact, proved by Liu, that the singular spheres are fixed components of the center and hence totally geodesic.

Corollary 1.4. Every manifold satisfying (1.1) is rationally elliptic. In particular,

Hodd(M;Q)=0,b2(M)=b4(M)=χ(M)−22≤2.H^{\mathrm{odd}}(M;\mathbb{Q}) = 0,\qquad b_2(M) = b_4(M) = \frac{\chi(M)-2}{2} \le2.

The possible real homotopy types are S6S^6, CP3\mathbb{C}\mathrm{P}^3, S2×S4S^2 \times S^4, S2×CP2S^2 \times\mathbb{C}\mathrm{P}^2, FWF_{\mathrm{W}}, and CP3#CP3\mathbb{C}\mathrm{P}^3\#\mathbb{C}\mathrm{P}^3.

This is a real-homotopy conclusion, not a replacement for integral reconstruction. In particular, the Eschenburg flag has the real homotopy type of CP3#CP3\mathbb{C}\mathrm{P}^3\#\mathbb{C}\mathrm{P}^3 but is not thereby identified with that connected sum.

Relation to Liu’s results

Liu’s corrected Theorem 1.1 gives the Euler characteristic, the principal-isotropy alternatives, restrictions on exceptional groups, and the bound on singular orbits. His Theorem 1.2 already classifies fixed-point-free SU(2)\mathrm{SU}(2) actions without exceptional orbits, and his Theorem 1.3 treats the exceptional-free mixed-boundary ball configuration. None of these results is claimed as new here.

The extensions are the weighted and integral refinements. For SO(3)\mathrm{SO}(3) with quotient S3S^3, the local orientation argument eliminates the disconnected-singular-isotropy alternative left in Liu’s Proposition 4.6. The global graph theorem of [37], combined with edge congruences and primitive normal Euler classes, gives the nine-family catalogue and the cyclic equivariant reconstruction. For SO(3)\mathrm{SO}(3) with quotient B3B^3, the sphere-bundle description suggested in Remark 4.5 is proved, the one-interior-orbit family is completely distinguished, the two-fixed-point/interior-orbit configuration is excluded by doubling, and the remaining exceptional configuration is recognized smoothly as CP3\mathbb{CP}^{3}. For SU(2), the new weight restriction and normal-line reconstruction extend Liu’s exceptional-free theorem to the exceptional cases in Figure 4.

The discussion of invariant product metrics is secondary. It establishes nonexistence within several broad geometric subclasses and identifies an action-specific boundary inequality for the SO(3) action. It neither removes S2×S4S^{2} \times S^{4} from the smooth candidate lists nor inserts unproved submersion or additional-symmetry assumptions into the classification arguments.

Common geometric and topological tools

Results used from the literature

We collect the precise inputs to separate the established theory from the reconstruction arguments below. The slice theorem and the classification of orbit bundles by normalizers are used in the form given by Bredon [5 Chapters IV–V]. The quotient of an isometric compact-group action on a manifold with sec⁡≥κ\sec\ge\kappa is an Alexandrov space with curvature at least κ\kappa; spaces of directions are the corresponding unit normal-sphere quotients [2].

Liu’s corrected Theorems 1.1–1.3, Theorem 3.1, and Proposition 4.9 give the following reductions [40]. The quotient is S3S^{3}, B3B^{3}, or B4B^{4}. Apart from the linear SO(3) actions on S(R3⊕Rtriv4)S(\mathbb{R}^{3} \oplus\mathbb{R}^{4}_{\mathrm{triv}}) and S(V5⊕Rtriv2)S(V_{5} \oplus\mathbb{R}^{2}_{\mathrm{triv}}), principal isotropy is trivial. The Euler characteristic is 2, 4, or 6. In cohomogeneity three every exceptional stratum has dimension at most one and the closure of each exceptional component meets a singular stratum. With quotient S3S^{3} there are at most three singular orbits. For SU(2) without fixed points, these are G/TG/T spheres, their union is the fixed set of the center, and exceptional isotropy is cyclic of odd order. Actions with nonempty fixed set, other than the finite-fixed-set SO(3) ball cases, are the linear cases identified by Liu using the low fixed-point-cohomogeneity classification of Grove–Kim and Grove–Searle [18, 21].

For a boundaryless quotient we use three results of Khalili Samani–Radeschi [37]: their Theorem A gives rational ellipticity; Proposition 5.1 gives the eighteen possible abstract quotient graphs; and their branched-cover and unknotting arguments, including Proposition 2.14 and Lemma 6.2, give the orbifold products needed in the cyclic reconstructions. The graph catalogue is a necessary list and is used as such. We do not assume that every listed graph is metrically realizable.

The boundary-soul theorem of Grove–Kim [18 Theorem 1.2], with its strata interpreted as in Liu’s Theorem 2.9, controls a positively curved quotient with boundary. In particular there is a unique farthest point from the boundary, and any isolated interior stratum is that point. We also use Petrunin’s metric doubling theorem [45 Theorem 2.1]. Doubling an Alexandrov space of curvature at least κ\kappa along its whole boundary preserves that lower bound. At a boundary point its new link is the double of the old link.

The extent inequality is used in the following form: four distinct points of a positively curved Alexandrov space cannot all have spaces of directions with three-extent at most π/3\pi/3 [19, 40]. The weighted Hopf quotients S3/S1S^{3}/S^{1} arising from two nonzero circle weights satisfy that bound [31]; the round Hopf quotient is S2(1/2)S^{2}(1/2). These general comparison results will be cited, not reproved.

For a closed simply connected six-manifold that is a union of two disk bundles, the rational-ellipticity theorem of DeVito–Galaz-García–Kerin requires b2≤3b_{2} \le3 [8 Theorem A]. This numerical hypothesis is essential in the general theorem. The decompositions used here have b2=1b_{2}=1 or 2 and therefore satisfy it.

Rational and integral recognition

A rationally elliptic space with positive Euler characteristic has no odd rational cohomology [13]. Together with Poincaré duality and Liu’s Euler-characteristic theorem, this proves, for every spherical quotient,

Hodd(M;Q)=0,b2=b4=(χ(M)−2)/2.(2)H^{\mathrm{odd}}(M;\mathbb{Q})=0,\qquad b_2=b_4=(\chi(M)-2)/2. \tag*{(2)}

Herrmann’s classification of rationally elliptic six-manifolds [30] gives the real homotopy types listed in Corollary 1.4. In rank two they can be read from the binary cubic v↦∫Mv3v\mapsto\int_M v^3: a double and a simple real root gives the S2×CP2S^2\times\mathbb{CP}^2 type; three distinct real roots give the flag type; and one real root gives the CP3#CP3\mathbb{CP}^3\mathbin{\#}\mathbb{CP}^3 type. The zero cubic and a pure cube do not give elliptic six-manifolds with b2=2b_2=2, b3=0b_3=0. These statements classify real, not integral, data.

For integral recognition we use the Wall–Jupp theorem [50, 33]. A closed simply connected oriented smooth six-manifold with torsion-free homology is determined up to oriented diffeomorphism by the isomorphism class of

(H2(M;Z), b3(M), μM, w2(M), pM),(3)\left(H^2(M;\mathbb{Z}),\ b_3(M),\ \mu_M,\ w_2(M),\ p_M\right), \tag*{(3)}

where

μM(x,y,z)=∫Mxyz,pM(x)=∫Mp1(M)x.\mu_M(x,y,z)=\int_M xyz,\qquad p_M(x)=\int_M p_1(M)x.

The isomorphism must respect the integral lattice and the reduction of w2w_2 modulo two. We use this classification only after proving torsion-freeness. More general six-dimensional classification results, such as [57], involve additional data; a rational cohomology computation is not a substitute for them. Finally, an integral homology six-sphere that is simply connected is a homotopy six-sphere, and Θ6=0\Theta_6=0 by Kervaire–Milnor [36]. This gives the ordinary smooth upgrade of Liu’s topological sphere case.

Weight normalization and singular-orbit tubes

Choose orientations so that the homogeneous weight-one line over G/T=S2G/T=S^2 has Chern number

dG={1,G=SU(2),2,G=SO(3).(4)d_G=\begin{cases} 1,&G=\mathrm{SU}(2),\\ 2,&G=\mathrm{SO}(3). \end{cases} \tag*{(4)}

Thus the tangent representation of TT on G/TG/T has weight 2/dG2/d_G. For positive coprime integers p,qp,q write

Up,qG=G×TD(Cp⊕Cq),Bp,qG=∂Up,qG=G×TS(Cp⊕Cq).(5)U^G_{p,q}=G\times_T D(\mathbb{C}_p\oplus\mathbb{C}_q),\qquad B^G_{p,q}=\partial U^G_{p,q}=G\times_T S(\mathbb{C}_p\oplus\mathbb{C}_q). \tag*{(5)}

The orbit space of the boundary is the spindle S2(p,q)S^2(p,q), with cone points of orders p,qp,q when these exceed one. The underlying oriented four-plane bundle over G/TG/T has clutching class dG(p+q) mod 2d_G(p+q)\bmod2 in π1(SO(4))=Z/2\pi_1(\mathrm{SO}(4))=\mathbb{Z}/2. Consequently, for every pair in the SO(3)\mathrm{SO}(3) discussion and for odd pairs in the SU(2)\mathrm{SU}(2) discussion,

Up,qG≅S2×D4,Bp,qG≅S2×S3(6)U^G_{p,q}\cong S^2\times D^4,\qquad B^G_{p,q}\cong S^2\times S^3 \tag*{(6)}

as ordinary manifolds. This does not make the equivariant boundary bundle trivial.

Lemma 2.1 (Equivariant boundary extension). For the tubes in (6), every GG-equivariant diffeomorphism of Bp,qGB^G_{p,q} is equivariantly isotopic to the restriction of a diffeomorphism of Up,qGU^G_{p,q}. In particular, the isotopy itself can be absorbed in a collar, so an attaching map to such a tube introduces no additional parameter once the other piece and its boundary identification have been fixed.

Proof. The boundary is a locally free GG-space over S2(p,q)S^2(p,q). An induced quotient diffeomorphism preserves the cone orders. An orientation-preserving orbifold diffeomorphism of this sphere is isotopic to the identity: remove invariant disks at its at most two cone points, use the marked version of the sphere isotopy theorem [48], and perform the isotopy in the cyclic disk charts. There are no two distinct equal nontrivial orders, since gcd⁡(p,q)=1\gcd(p,q)=1. Lifting this isotopy with an invariant connection reduces the problem to an equivariant map over the identity.

Such maps are sections of the adjoint orbibundle. In a chart with isotropy CmC_m, the value at the chart center lies in its centralizer. For G=SU(2)G=\mathrm{SU}(2), every nontrivial mm here is odd and its centralizer is a connected circle. Move these values to the identity through their centralizers. Obstruction theory on the remaining two-dimensional orbifold charts now reduces the section to the identity: π1(SU(2))=π2(SU(2))=0\pi_1(\mathrm{SU}(2))=\pi_2(\mathrm{SU}(2))=0.

For G=SO(3)G=\mathrm{SO}(3) one must account for the disconnected centralizer of a half-turn. Consider the fiberwise double cover of the adjoint bundle with fiber SU(2)\mathrm{SU}(2). The obstruction to lifting a section is a class in

Horb1(S2(p,q);Z/2)=0.H^1_{\mathrm{orb}}(S^2(p,q);\mathbb{Z}/2)=0.

Indeed the orbifold fundamental group has presentation ⟨a,b∣ap=bq=ab=1⟩\langle a,b\mid a^p=b^q=ab=1\rangle and is cyclic of order gcd⁡(p,q)=1\gcd(p,q)=1. This also shows why a value in a disconnected half-turn centralizer cannot be chosen independently at a cone point. A lifted section has its values in the centralizers of the lifted cyclic subgroups of SU(2)\mathrm{SU}(2), which are connected. The preceding contraction and the same vanishing homotopy groups apply. This is the usual obstruction-theoretic description of orbibundle gauge sections; see [1] for the orbifold bundle language.

It remains to treat quotient orientation reversal. Let w∈NG(T)∖Tw\in N_G(T)\setminus T and let cc be complex conjugation on both weight summands. Since c(tv)=t−1c(v)c(tv)=t^{-1}c(v) and wtw−1=t−1wtw^{-1}=t^{-1}, the map

[g,v]⟼[gw,c(v)][g,v]\longmapsto[gw,c(v)]

is a well-defined equivariant diffeomorphism of the whole tube. It reverses the orientation of its spindle quotient. Composing with it reduces the remaining case to the orientation-preserving one. The isotopies constructed above extend across an invariant collar. Smoothness at a cyclic chart center follows by working equivariantly in the covering disk throughout. □

The lifting argument is needed only for the nonfree spindle boundary. For a principal GG-bundle over S2S^2, an automorphism over the identity is a section of the ordinary adjoint bundle. For SO(3)\mathrm{SO}(3) the lift obstruction is in H1(S2;Z/2)=0H^1(S^2;\mathbb{Z}/2)=0, and π2(SU(2))=0\pi_2(\mathrm{SU}(2))=0 again makes the gauge section null-homotopic. For SU(2)\mathrm{SU}(2) this is immediate. We will use this free-boundary version in the ball cases.

A closed exceptional interval and its normal bundle

Suppose a closed exceptional interval of order mm has two circle-isotropy endpoints and no finite branching vertex. Its full inverse image, including the endpoints, is the smooth cohomogeneity-one four-manifold

Am=G×TS(R⊕Cm)=P(1⊕LdGm)⟶S2.(7)A_m=G\times_T S(\mathbb{R}\oplus\mathbb{C}_m)=\mathbb{P}(1\oplus L^{d_Gm})\longrightarrow S^2. \tag*{(7)}

Here the local axis planes in the two endpoint slices provide the smooth caps. Denote the fiber class by ff and the two orbit sections, with their G/TG/T orientations, by S−,S+S_-,S_+. With a compatible choice of labeling,

[S+]−[S−]=dGm[f].(8)[S_+]-[S_-]=d_Gm[f]. \tag*{(8)}

Lemma 2.2 (Normal characters). An oriented equivariant normal two-plane bundle ν\nu of AmA_m is determined by its oriented endpoint characters u,v∈Zu,v\in\mathbb{Z}. They must satisfy v≡u(modm)v\equiv u\pmod m, and its Euler class obeys

⟨e(ν),[S−]⟩=dGu,⟨e(ν),[S+]⟩=dGv,⟨e(ν),[f]⟩=v−um.(9)\langle e(\nu),[S_-]\rangle=d_Gu,\qquad\langle e(\nu),[S_+]\rangle=d_Gv,\qquad\langle e(\nu),[f]\rangle=\frac{v-u}{m}. \tag*{(9)}

The unit circle bundle of ν\nu is simply connected if and only if

gcd⁡(dGu,v−um)=1.(10)\gcd\left(d_Gu,\frac{v-u}{m}\right)=1. \tag*{(10)}

Proof. At a principal orbit G/CmG/C_m the two endpoint representations restrict to the same complex character, which is exactly u≡v(modm)u\equiv v\pmod m. The normal bundle can be constructed on the two cohomogeneity-one disk bundles over the sections using these characters. On their common principal orbit any two complex-linear intertwiners differ by an element of C∗\mathbb{C}^*, a connected group. Thus no additional clutching component remains. The restrictions to the sections have Chern numbers dGud_Gu, dGvd_Gv, and (2.7) gives the fiber evaluation. Conversely these constructions realize every compatible pair. Since AmA_m is simply connected, the homotopy sequence of its oriented circle bundle identifies its fundamental group with Z\mathbb{Z} modulo the image of evaluation by e(ν)e(\nu). The section and fiber form an integral basis of H2(Am)H_2(A_m), proving (10). □

A real transverse representation records a character only up to sign. The signs in Lemma 2.2, in contrast, are fixed by an orientation of the normal bundle. This distinction is responsible for several restrictions that are invisible in an unsigned orbit diagram.

Localization at the singular spheres

For a spherical quotient with connected singular isotropy, let S1,…,SrS_1,\ldots,S_r be the singular spheres, with slice weights (pi,qi)(p_i,q_i). Give each sphere the orientation from an identification with SU(2)/T\mathrm{SU}(2)/T. For x∈H2(M;Z)x \in H^2(M;\mathbb{Z}) put ni(x)=∫Sixn_i(x)=\int_{S_i}x. The SU(2)\mathrm{SU}(2) action used here is the given action when G=SU(2)G=\mathrm{SU}(2) and the double-cover action when G=SO(3)G=\mathrm{SO}(3).

Proposition 2.3. There are signs εi∈{1,−1}\varepsilon_i \in\{1,-1\} determined by the tangent orientations such that

0=∑iεini(x)piqi.(11)0=\sum_i\frac{\varepsilon_i n_i(x)}{p_iq_i}. \tag*{(11)}
μM(x,y,z)=1dG2∑iεini(x)ni(y)ni(z)piqi.(12)\mu_M(x,y,z)=\frac{1}{d_G^2}\sum_i\frac{\varepsilon_i n_i(x)n_i(y)n_i(z)}{p_iq_i}. \tag*{(12)}
pM(x)=∑iεini(x)(4/dG2+pi2+qi2)piqi.(13)p_M(x)=\sum_i\frac{\varepsilon_i n_i(x)(4/d_G^2+p_i^2+q_i^2)}{p_iq_i}. \tag*{(13)}

The rational evaluation map H2(M;Q)→QrH^2(M;\mathbb{Q})\to\mathbb{Q}^r is injective. When χ(M)=2r\chi(M)=2r, its image is exactly the hyperplane (11).

Proof. We apply the equivariant localization theorem [3, 4]; only its specialization needs explanation. Since Hj(BSU(2);Z)=0H^j(B\mathrm{SU}(2);\mathbb{Z})=0 for 1≤j≤31\leq j\leq3, every degree-two class has a unique SU(2)\mathrm{SU}(2)-equivariant lift. At the two circle-fixed points of SiS_i, the lift restricts to ni(x)tn_i(x)t and −ni(x)t-n_i(x)t, where tt generates H2(BT)H^2(BT). The tangent weights for this circle are

2, dGpi, dGqi.2,\ d_Gp_i,\ d_Gq_i.

Their Euler products are 2εidG2piqit32\varepsilon_i d_G^2p_iq_it^3 and its negative, while the equivariant first Pontryagin class is (4+dG2pi2+dG2qi2)t2(4+d_G^2p_i^2+d_G^2q_i^2)t^2 at either point. Applying localization to xx, to xyzxyz, and to p1xp_1x gives (11)–(13). The integration of the degree-two lift is zero for dimensional reasons.

Equation (2) gives equivariant formality over Q\mathbb{Q}. Restriction of equivariant cohomology to the circle-fixed set is therefore injective. If all ni(x)n_i(x) vanish, the unique equivariant lift vanishes and so does xx. The image has dimension b2=r−1b_2=r-1, which proves the hyperplane assertion. □

For SO(3)\mathrm{SO}(3) the factor 1/41/4 in (12) is indispensable. The proposition determines rational forms on an evaluation space. It does not identify the integral evaluation lattice or establish torsion-freeness of homology.

SO(3) actions with orbit space S3

Throughout this section G=SO(3)G=\mathrm{SO}(3), X≅S3X\cong S^3, and (1) holds. Principal isotropy is trivial by Liu. We first determine the local possibilities, then use the global graph theorem, and finally recover the smooth manifold in the cyclic cases.

Connected singular isotropy and faithful finite slices

Lemma 3.1. Every singular isotropy group is T=SO⁡(2)T = \operatorname{SO}(2). Its normal slice is Cp⊕Cq\mathbb{C}_p \oplus\mathbb{C}_q with nonzero coprime weights. If rr denotes the number of singular orbits, then

1≤r≤3,χ(M)=2r.1 \le r \le3,\qquad\chi(M) = 2r.

Proof. Fixed points would lie in the boundary of the quotient, by the six-dimensional fixed-point representations listed by Liu. A remaining singular isotropy group is TT or its normalizer O(2)O(2). At an interior isolated singular orbit both circle weights on the four-dimensional slice are nonzero. An element of O(2)∖TO(2) \setminus T conjugates the circle to its inverse. On the slice it is anti-complex-linear on two complex dimensions, hence has positive real determinant, including when the two weights have equal absolute value. On the tangent plane of G/O(2)=RP2G/O(2) = \mathbb{R}P^2 it reverses orientation. It would therefore reverse the orientation of MM. This is impossible for an element of the connected group GG.

Thus every singular orbit is G/TG/T. Slice effectiveness and trivial principal isotropy give gcd⁡(p,q)=1\gcd(p,q) = 1, while a zero weight would produce a boundary face. Every singular sphere contains two TT-fixed points and no principal or exceptional orbit contains any. The Euler formula and singular-orbit bound of Liu now give the result. □

The determinant observation is the boundaryless counterpart of the orientation analysis already used in Liu’s ball-quotient proof; it removes the disconnected-isotropy alternative from his Proposition 4.6.

At an exceptional orbit G/ΓG/\Gamma, the slice has real dimension three. Its representation is faithful, since a nontrivial slice kernel would occur in principal isotropy nearby. It preserves orientation: both the adjoint action on the orbit tangent space and the ambient GG action do so. The finite rotation groups therefore give the following local incidence data [5, 40]:

Finite isotropyIncident edge orders in the quotient slice
DmD_m2, 2, mm
A4A_42, 3, 3
S4S_42, 3, 4
A5A_52, 3, 5

Table 6.

A cyclic edge has slice R⊕Cs\mathbb{R} \oplus\mathbb{C}_s for a generator of CmC_m, with gcd⁡(s,m)=1\gcd(s,m) = 1. At a TT vertex the two slice weights record the incident orders greater than one, so its valency is at most two. The representation of a finite group, not just its isomorphism type, remains part of the data. In particular the two faithful three-dimensional representations of A5A_5, relative to a fixed embedding, can have order-five transverse weights ±1\pm1 and ±2\pm2.

Lemma 3.2 (Edge congruence). For an order- mm edge, the endpoint transverse characters satisfy u≡±v(modm)u \equiv\pm v \pmod m. The only ambiguity in comparing the characters is inversion, not multiplication by an arbitrary unit of Z/m\mathbb{Z}/m.

Proof. Along a connected open edge, its normal slice representation is locally constant after identifying its isotropy subgroup. Conjugation in SO⁡(3)\operatorname{SO}(3) between two identifications of a rotation subgroup is induced by an element of its normalizer and acts on the rotation circle by identity or inversion. Restricting the endpoint slice to this subgroup gives the asserted congruence. □

The nine necessary weighted graphs

Table 1 gives the full abstract list. Filled vertices in Figure 1 have circle isotropy; hollow vertices have noncyclic finite isotropy. The list does not discard the representation qualification in Table 2.

Proposition 3.3. The graph Σ\Sigma, after forgetting its embedding in S3S^3, belongs to exactly one family in Table 1, up to graph isomorphism and the indicated parameter symmetries.

Proof. We start with the eighteen graphs of [37 Proposition 5.1]. The specialization is explicit: graphs 1, 2, 3 there are the cases with no edges; graphs 4, 5, 6, 8, 9 give II, IV, VII, III, V; graphs 7, 18 contain a one-vertex loop; graphs 10, 12, 13, 15, 16 have a three-valent non-orbifold vertex; graph 11 gives VIII, graph 14 the subdivided theta, and graph 17 the point-plus-interval. A non-orbifold vertex here has circle isotropy and valency at most two, so the five three-valent cases are impossible. A one-vertex loop of order m>1m > 1 would force the slice pair (m,m)(m,m), contrary to coprimality.

For a path order its labels a<ba < b. At the order-bb edge, the endpoint transverse characters are 1,a1,a, so a≡±1(modb)a \equiv\pm1 \pmod b. With a≥2a \ge2, this gives a=b−1a=b-1. For a genuine triangle let a<b<ca < b < c be its labels. Along the cc edge, a≡±b(modc)a \equiv\pm b \pmod c. The positive sign is impossible and the negative sign gives c=a+bc=a+b. These are IV and V.

The global graph theorem gives an order-two edge on the cycle of a lollipop. Write its other cycle order as kk and its stem order as ℓ\ell. Coprimality at the cycle’s circle vertex forces kk odd. Substitution of (2,k,ℓ)(2,k,\ell) in the four finite-slice triples gives precisely Table 2. Comparing the transverse character 2 on the order-kk edge and the character 1 on the stem gives the last column. The same comparison at the order-mm leaf of a YY forces its finite-slice weight to be ±1\pm1.

For the theta reduction we use the exceptional-track obstruction of [37 Lemma 2.10]. A complete quotient geodesic track that continues through finite vertices cannot avoid every singular vertex. In a finite slice, continuation pairs antipodal axis directions. At DmD_m, the two order-two arms are interchanged when mm is odd, while all arms reflect individually when mm is even. At A4A_4 the two order-three arms are interchanged and the order-two arm reflects. At $S_4 and A5A_5 every arm reflects. These statements follow directly from the rotation-axis orbits on the unit two-sphere.

Let the two direct edges of a subdivided theta have orders r,qr,q. If both are order two, they form a finite-only track, whether the endpoints interchange or reflect them. If they have different orders, a direct arm reflected at both endpoints likewise gives a forbidden track. The only available interchange mechanisms are the order-two pair at an odd dihedral vertex and the order-three pair at a tetrahedral vertex. Both must occur. Thus the direct orders are 2, 3, the finite vertices are D3D_3, A4A_4, and the remaining arms have orders 2, 3 and meet at the sole circle vertex. This gives IX.

Two topological restrictions complete the argument. If there are no edges, all slices are (1,1)(1,1). Remove rr tubes U1,1GU_{1,1}^{G}. The remainder is a principal SO(3)\mathrm{SO}(3) bundle over S3S^3 with rr open balls removed. Each boundary restriction has nonzero w2w_2, because its total space S2×S3S^2 \times S^3 is simply connected. The sum of the boundary sphere classes is zero modulo two, so evaluating the bundle’s w2w_2 gives r≡0(mod2)r \equiv0 \pmod{2}. Since 1≤r≤31 \le r \le3, only I remains.

Finally suppose an order-mm interval is disjoint from an isolated (1,1)(1,1) vertex. A tube about its inverse image AmA_m and the isolated-orbit tube have a common boundary diffeomorphic to S2×S3S^2 \times S^3. The normal characters of AmA_m have absolute values one. Set u=1u=1, v=±1v=\pm1 in Lemma 2.2. The choice v=1v=1 has Euler evaluations 2, 0 and a non-simply-connected unit circle bundle. Hence v=−1v=-1 and m∣2m \mid2. Since m>1m>1, m=2m=2, which is VI. The decomposition used here follows by removing the two disjoint quotient balls and lifting the interval direction on their free complement S2×[0,1]S^2 \times[0,1]. □

The nine necessary abstract weighted graphs I–IX, showing circle-isotropy and noncyclic finite-isotropy vertices

Figure 1. The nine necessary abstract weighted graphs. Filled vertices have circle isotropy and hollow vertices have noncyclic finite isotropy. The picture specifies incidence, not a classification of spatial embeddings or positive-curvature realizations.

Smooth reconstruction in the cyclic cases.

For I–III, remove the two singular-orbit tubes. In III the bigon cycle is unknotted by the branched-cover argument in [37 Section 6, graph 8]. In II a single tame arc with its endpoints free in S3S^3 can be straightened; in I there is no arc. The complement, with its quotient smooth structure, is S2(p,q)×[0,1]S^2(p,q)\times[0,1], using [37 Lemma 6.2]. An invariant connection lifts the interval direction and identifies the corresponding GG-manifold with a product. Therefore

M=Up,qG∪φUp,qG.M=U^G_{p,q}\cup_{\varphi}U^G_{p,q}.

Lemma 2.1 absorbs φ\varphi. The standard double is Np,qGN^G_{p,q}. Its underlying oriented five-plane bundle over S2S^2 has w2=2(p+q)mod⁡2=0w_2=2(p+q)\mathbin{\operatorname{mod}}2=0, so it is trivial. This proves Np,qG≅S2×S4N^G_{p,q}\cong S^2\times S^4.

For VI the argument at the end of Proposition 3.3 determines the normal line on A2A_2 uniquely, up to simultaneous reversal of the normal orientation and exchange of its sections. Its Euler evaluations are 2,−12,-1 on a section and a fiber. The last gluing is absorbed by U1,1GU^G_{1,1}. The projective model P1,1P_{1,1} realizes these data, proving uniqueness equivariantly.

For IV or V choose an edge of order a≥2a\geq2 whose endpoint transverse absolute weights are b,a+bb,a+b, with gcd⁡(a,b)=1\gcd(a,b)=1. The path case is b=1b=1; in the triangle case both a,b>1a,b>1. The third singular vertex has slice (b,a+b)(b,a+b). In the triangle case the cycle is unknotted by [37 Section 6, graph 9]; the orbifold product lemma then identifies the intervening region. Consequently

M=D(νAa)∪S2×S3Ub,a+bG.(14)M=D(\nu A_a)\cup_{S^2\times S^3}U^G_{b,a+b}. \tag*{(14)}

Orient the normal line so that its first character is u=bu=b. The second is v=±(a+b)v=\pm(a+b). The positive choice is compatible and has fiber Euler number one. The negative choice requires a∣2ba\mid2b, hence a=2a=2. Then bb is odd, and both Euler evaluations 2b2b and −(b+1)-(b+1) are even. This contradicts the simple connectedness of the common boundary in (14). Thus the positive choice is forced. Lemmas 2.2 and 2.1 now determine the normal line and remove the final gluing ambiguity.

For the induced projective action

z[z0:z1:z2]=[z0:zaz1:za+bz2],Pa,b=G×TCP2,(15)z[z_0:z_1:z_2]=[z_0:z^a z_1:z^{a+b}z_2],\qquad P_{a,b}=G\times_T\mathbb{CP}^2, \tag*{(15)}

the three coordinate fixed points have slice pairs (a,a+b)(a,a+b), (a,b)(a,b), and (b,a+b)(b,a+b). Coordinate projective lines give the three possible edges, with labels one omitted. Thus Pa,bP_{a,b} realizes precisely the forced gluing data. This proves the equivariant assertions of Theorem 1.1 for I–VI. The uniqueness argument uses the primitive normal Euler class and the boundary extension; it does not follow from the unsigned graph alone.

Let u=c1(L)u=c_1(L) on the base and h=c1(O(1))h=c_1(\mathcal{O}(1)) on the projectivization, where P\mathbb{P} parametrizes lines. Put c=2(2a+b)c=2(2a+b). The projective-bundle formula and the vertical Euler sequence [41] give

H∗(Pa,b;Z)=Z[u,h]/(u2,h3+cuh2),∣u∣=∣h∣=2,∫Pa,buh2=1.(16)H^*(P_{a,b};\mathbb{Z})=\mathbb{Z}[u,h]/(u^2,h^3+cuh^2),\qquad|u|=|h|=2,\qquad\int_{P_{a,b}}uh^2=1. \tag*{(16)}
Hodd(Pa,b;Z)=0,w2(Pa,b)=hˉ,p1(Pa,b)=3h2+2cuh.(17)H^{\mathrm{odd}}(P_{a,b};\mathbb{Z})=0,\qquad w_2(P_{a,b})=\bar{h},\qquad p_1(P_{a,b})=3h^2+2cuh. \tag*{(17)}

Indeed, the stable vertical tangent roots are h,h+2au,h+2(a+b)uh,h+2au,h+2(a+b)u. Summing their squares gives (17). Linear projective bundles over S2S^2 are classified by π1(PU(3))=Z/3\pi_1(PU(3))=\mathbb{Z}/3; their clutching invariant here is

c≡a−b(mod3).c\equiv a-b\pmod{3}.

The two nonzero classes are interchanged by complex conjugation. To see that neither is diffeomorphic to the zero class, note that the primitive square-zero class in (16) is uu up to sign. Changing the other generator to ±h+ju\pm h+ju changes cc only by sign and a multiple of three. The two claimed ordinary diffeomorphism types therefore are distinct. All these cyclic models admit invariant nonnegative curvature, obtained from a bi-invariant metric on GG and a round or Fubini–Study metric by the free diagonal TT quotient; strict positivity is not a consequence of this construction.

A smooth family with dihedral branching.

Proposition 3.4. For each odd k≥3k \ge3 there is a smooth simply connected SO(3)\mathrm{SO}(3)-manifold with type-VIII graph (Dk;2,k;2)(D_k; 2,k;2) whose underlying manifold is

En=S(ξn)⟶S4,n=k2−18,p1(ξn)=4nu.E_n=S(\xi_n)\longrightarrow S^4,\qquad n=\frac{k^2-1}{8},\qquad p_1(\xi_n)=4nu.

This is a smooth realization and does not assert positive curvature for every kk.

Proof. Use the cohomogeneity-one action on S4⊂Sym⁡02(R3)S^4\subset\operatorname{Sym}_0^2(\mathbb{R}^3), with principal isotropy D2D_2 and singular isotropy K−,K+≅O(2)K_-,K_+\cong O(2) about different coordinate axes. Define homomorphisms into a second SO(3)\mathrm{SO}(3) by multiplying the circle angle by kk on K−K_- and using the ordinary inclusion on K+K_+. Send the reflection in each normalizer to the corresponding half-turn. Because kk is odd, the homomorphisms agree on D2D_2. The equivariant principal-bundle construction of Grove–Ziller [26 Proposition 1.6] gives a principal SO(3)\mathrm{SO}(3) bundle over S4S^4 with a commuting lifted GG action. Take its associated two-sphere bundle.

Over the principal base orbit the three axes of D2D_2 give three order-two arcs. At the weight-kk end the fiber poles produce a circle orbit with slice (2,k)(2,k), while the equator produces DkD_k. At the weight-one end the poles give slice (1,2)(1,2) and the equator has cyclic isotropy. These arcs assemble into the required lollipop; at its DkD_k vertex the order-kk transverse character is ±2\pm2.

At the two fixed points of a rotation circle on the base S4S^4, the tangent weights are (1,2)(1,2) with opposite Euler signs. The associated three-plane fiber weights are kk and 1. Localization gives

∫S4p1(ξn)=k2−12=4n.\int_{S^4}p_1(\xi_n)=\frac{k^2-1}{2}=4n.

The total sphere bundle is simply connected by the homotopy sequence. This proves the assertion, and its integral invariants are computed in (24) below. □

What the branching graphs determine.

In type VIII let xx generate H2(M;Z)≅ZH^2(M;\mathbb{Z})\cong\mathbb{Z}, and set g0=gcd⁡(ℓ,2k)g_0=\gcd(\ell,2k). The evaluation relation (11) permits compatible orientations and a positive integer tt such that

nleaf(x)=ℓtg0,ncycle(x)=2ktg0,n_{\mathrm{leaf}}(x)=\frac{\ell t}{g_0},\qquad n_{\mathrm{cycle}}(x)=\frac{2kt}{g_0},

with orientation signs −-, ++. Equations (12)–(13) yield

d:=∫Mx3=(4k2−ℓ2)t34g03,P:=∫Mp1(M)x=(k2+3−ℓ2)tg0.(18)d:=\int_M x^3=\frac{(4k^2-\ell^2)t^3}{4g_0^3},\qquad P:=\int_M p_1(M)x=\frac{(k^2+3-\ell^2)t}{g_0}. \tag*{(18)}

Since 2k>ℓ2k>\ell in every permitted row, d>0d>0. Thus these manifolds have rational cohomology Q[x]/(x4)\mathbb{Q}[x]/(x^4), not the rational ring of S2×S4S^2\times S^4.

A primitive degree-two class need not have a primitive vector of evaluations on the singular spheres. Thus tt cannot be set equal to one without an integral lattice computation. Proposition 3.4 realizes t=1t=1 in the dihedral row. If homology is torsion-free, Wall–Jupp further requires

P≡4d(mod24),w2=0,d≡0(mod2),P≡d(mod48),w2=xˉ.(19)\begin{aligned} P&\equiv4d &&\pmod{24}, & w_2&=0,\\ d&\equiv0 &&\pmod{2}, & P&\equiv d &&\pmod{48}, & w_2&=\bar{x}. \tag*{(19)} \end{aligned}

These congruences are tests, not proofs, that an arbitrary listed integer is realized.

For a type-VII YY, absorb the orientation signs into the evaluations u,v,wu,v,w on the two order-two leaves and the order-mm leaf. Then

u2+v2+wm=0,w=−m2(u+v).\frac{u}{2}+\frac{v}{2}+\frac{w}{m}=0,\qquad w=-\frac{m}{2}(u+v).
Finite vertexParameterddPP
DkD_k, kk oddt≥1t \ge1(k2−1)t3/8(k^2-1)t^3/8(k2−1)t/2(k^2-1)t/2
A4A_4t=2ht=2h, h≥1h \ge12h32h^32h2h
S4S_4t=2ht=2h, h≥1h \ge15h35h^3−4h-4h
A5A_5, (k,ℓ)=(3,5)(k,\ell)=(3,5)t=2ht=2h, h≥1h \ge122h322h^3−26h-26h
A5A_5, (k,ℓ)=(5,3)(k,\ell)=(5,3)t=2ht=2h, h≥1h \ge1182h3182h^338h38h

Table 3. Necessary characteristic numbers for lollipop graphs. Integrality forces tt even in the final four rows. These are arithmetic restrictions, not an existence list.

On this rational plane, localization gives

μ(x,x,x)=u3+v38−m2(u+v)332,(20)\mu(x,x,x)=\frac{u^3+v^3}{8}-\frac{m^2(u+v)^3}{32}, \tag*{(20)}
pM(x)=4−m22(u+v).(21)p_M(x)=\frac{4-m^2}{2}(u+v). \tag*{(21)}

In the coordinates s=u+vs=u+v, d=u−vd=u-v, the cubic is

132s(3d2−(m2−1)s2).\frac{1}{32}s\left(3d^2-(m^2-1)s^2\right).

It has three distinct real linear factors for every m≥2m \ge2. Hence these manifolds have the real homotopy type of FWF_W, although the rational Pontryagin class vanishes only when m=2m=2. In particular m>2m>2 cannot give the ordinary flag manifold. The integral evaluation lattice, torsion, and w2w_2 are not determined by this rational calculation.

For IX, r=1r=1 gives the rational cohomology of S6S^6. If an integral gluing computation proves that homology is torsion-free, the manifold is diffeomorphic to S6S^6 by [36]. This conditional recognition does not suppress possible torsion.

Three known positively curved actions check the catalogue. The irreducible action on S(H3(R3))S(H_3(\mathbb{R}^3)) has one circle orbit with slice (2,3)(2,3), so its graph is IX; Liu’s harmonic cubic xyzxyz supplies the A4A_4 orbit. The projectivized irreducible action on CP3=P(Sym⁡3C2)\mathbb{C}P^3=\mathbb{P}(\operatorname{Sym}^3\mathbb{C}^2) has slices (2,3)(2,3) and (1,2)(1,2) and graph VIII with D3D_3. The left SO(3)\mathrm{SO}(3) action on FWF_W has graph VII with m=2m=2: its finite stabilizers are intersections with tori, so the noncyclic one is the abelian group D2D_2 [40 Examples 5.2–5.4]. These identifications verify the examples; they do not prove that a branching graph determines a unique manifold.

SO(3) actions with orbit space B3

Let G=SO(3)G=\mathrm{SO}(3) and X≅B3X\cong B^3. The action with nontrivial principal isotropy is already a linear action on S6S^6, by Liu’s Theorem 3.2. We therefore assume principal isotropy is trivial and the fixed set is finite or empty. Liu’s Theorem 4.2 gives a boundary face with TT isotropy, at most one interior singular orbit, and the alternatives in his fig:2. We retain the letters of that figure, but describe every case in words in Table 4 so that no external illustration is needed.

CaseBoundary and interiorSmooth conclusionCurvature conclusion
(a)Nontrivial principal isotropy; fixed circle and two faces.Linear S6S^6; Liu.Realized on the round sphere.
(b)Uniform TT boundary; principal interior.En=S(ξn)→S4E_n=S(\xi_n)\to S^4, n∈Zn\in\mathbb{Z}.No parameter bound from the present argument.
(c)Uniform TT boundary; one interior TT orbit.MkM_k, kk odd, distinguished by ∣k∣|k|.Smoothly realizable; strict positivity not decided.
(d)Two boundary fixed points; no interior singular orbit.Standard S6S^6.Realized by the diagonal linear action.
(e)Two boundary fixed points; one interior TT orbit.Smoothly realized on a complex quadric.Impossible under (1.1).
(f)One boundary fixed point, one boundary O(2)O(2) orbit, one interior TT orbit, and a C2C_2 interval.CP3\mathbb{C}P^3.Realized by the real projective block action.

Table 4. The SO(3)SO(3) ball-quotient alternatives, with Liu’s Figure 2 lettering. The topology of the entire boundary, not just a boundary face, is used.

The full stratification and the boundary normal line

At a point of a boundary face the slice is Rtriv2⊕C1\mathbb{R}^2_{\mathrm{triv}}\oplus\mathbb{C}_1. At an isolated GG-fixed point it is R3⊕R3\mathbb{R}^3\oplus\mathbb{R}^3. Neither slice contains exceptional orbits. At the isolated boundary O(2)\mathrm{O}(2) orbit in (f), the slice is

Rdet⊕Rdet⊕Cstd.(22)\mathbb{R}_{\mathrm{det}}\oplus\mathbb{R}_{\mathrm{det}}\oplus\mathbb{C}_{\mathrm{std}}. \tag*{(22)}

It has exactly one exceptional germ, of order two, arising from a reflection in O(2)\mathrm{O}(2) rather than from its identity component. These are the corrected slice alternatives in Liu’s Theorem 2.11 and the proof of his Theorem 4.2 [40].

Lemma 4.1. In cases (b)–(f), no exceptional stratum occurs other than the interval displayed in (f). The interior slice is (1,1)(1,1) in (c) and (e), and (1,2)(1,2) in (f).

Proof. An interior noncyclic finite isotropy group would give an isolated interior stratum. By the boundary-soul theorem it must be the soul. In cases with an interior singular orbit the soul is already that orbit, so such a finite stratum is impossible. In case (b) there are no eligible boundary endpoints for any exceptional edge, and an exceptional component disjoint from singular strata is excluded by Liu’s Proposition 3.6. The same argument applies to (d), since its boundary fixed-point slices have no exceptional germs.

In (c) and (e) all exceptional endpoints would have to be the sole interior singular orbit. With no branching vertex available, an exceptional edge would then close to a loop at that orbit. Both slice weights would be its order, contradicting coprimality. Thus there are no exceptional edges and the two nonzero slice weights have absolute value one. In (f), the boundary normalizer orbit supplies exactly one order-two germ. It must end at the soul, since a face or a fixed point cannot be an endpoint and there is no other boundary normalizer orbit. A second edge or an additional component would have no possible endpoint except a prohibited loop at the soul. The remaining interior weight is therefore one. □

When the boundary has a single orbit type, let F=π−1(∂X)F=\pi^{-1}(\partial X). It is a smooth codimension-two submanifold. Its orbit bundle over S2S^2 has structure group NG(T)/T=Z/2N_G(T)/T=\mathbb{Z}/2 and is consequently trivial. Fix

F=So2×Sb2,e(νF)=2a+kb,(23)F=S^2_o\times S^2_b,\qquad e(\nu_F)=2a+kb, \tag*{(23)}

where a,ba,b generate the cohomology of the orbit and quotient factors, respectively. The coefficient two is forced by the weight-one normal character on G/TG/T. The integer kk is a genuine normal-bundle invariant. If η\eta is the homogeneous complex line over So2S^2_o with c1(η)=2ac_1(\eta)=2a and c1(L)=bc_1(L)=b, then the equivariant normal line is

νk=η⊠Lk.\nu_k=\eta\boxtimes L^k.

This also constructs the full boundary tube Vk=D(νk)V_k=D(\nu_k).

Uniform boundary with principal interior: the sphere bundles.

Proposition 4.2. In case (b), k=2nk=2n is even, and the underlying manifold is En=S(ξn)→S4E_n=S(\xi_n)\to S^4 with p1(ξn)=4nup_1(\xi_n)=4nu. Its integral data, in compatible generators x∈H2(M)x\in H^2(M) and y∈H4(M)y\in H^4(M), are

Hodd(M;Z)=0,H2(M)=H4(M)=Z,∫Mxy=1,x2=ny,H^{\mathrm{odd}}(M;\mathbb{Z})=0,\qquad H^2(M)=H^4(M)=\mathbb{Z},\qquad\int_M xy=1,\qquad x^2=ny,
w2(M)=0,p1(M)=4ny.(24)w_2(M)=0,\qquad p_1(M)=4ny. \tag*{(24)}

Every integer nn is smoothly realizable in this stratum, and the ordinary diffeomorphism type is determined by ∣n∣|n|.

Proof. After removing the boundary collar, the principal region is U=G×D3U=G\times D^3. Thus

M=U∪WVk,W=S(νk)≅G×S2.M=U\cup_W V_k,\qquad W=S(\nu_k)\cong G\times S^2.

For a circle bundle over the simply connected base FF, the fundamental group of its total space has order equal to the divisibility of its Euler class. Hence π1(W)=Z/2\pi_1(W)=\mathbb{Z}/2 forces k=2nk=2n. Put c=a+nbc=a+nb, so e(νk)=2ce(\nu_k)=2c.

The Gysin sequence gives H2(W)=Z⊕Z/2H^2(W)=\mathbb{Z}\oplus\mathbb{Z}/2 and H4(W)=Z/2H^4(W)=\mathbb{Z}/2. The torsion generator from H2(U)=Z/2H^2(U)=\mathbb{Z}/2 restricts to the class of cc modulo 2c2c. The restriction H3(U)→H3(W)H^3(U)\to H^3(W) is an isomorphism, since it restricts as an equivariant degree-one map on a principal GG orbit. Mayer–Vietoris now gives torsion-free cohomology in degrees 0, 2, 4, 6, one generator in each degree, and zero odd cohomology. Generators can be chosen with

x∣F=c,y∣F=2ab,∫Mxy=1.x|_F=c,\qquad y|_F=2ab,\qquad\int_Mxy=1.

For the last normalization, PD⁡[F]=2x\operatorname{PD}[F]=2x and ∫Fy∣F=2\int_Fy|_F=2 imply ∫Mxy=1\int_Mxy=1. Restriction to FF is injective in degree four. It follows from c2=2nabc^2=2nab that x2=nyx^2=ny.

Since TM∣F=TF⊕νkTM|_F=TF\oplus\nu_k, FF is spin, and e(νk)=2ce(\nu_k)=2c, restriction gives w2(M)∣F=0w_2(M)|_F=0. Restriction in degree two modulo two is injective, so w2(M)=0w_2(M)=0. Moreover p1(TF)=0p_1(TF)=0 and

p1(M)∣F=e(νk)2=8nab=4n y∣F,p_1(M)|_F=e(\nu_k)^2=8nab=4n\,y|_F,

which proves (24).

Oriented three-plane bundles over S4S^4 are classified by an integer normalized by p1(ξn)=4nup_1(\xi_n)=4nu [41, 32]. Their sphere bundles have exactly (24): lift the structure group to SU⁡(2)\operatorname{SU}(2) and use its rank-two complex projectivization, or use the vertical Euler class 2x2x and the stable tangent splitting. Wall–Jupp identifies MM with EnE_n. The cubic and its integral divisibility distinguish ∣n∣|n|; simultaneously changing x,yx,y to −x,−y-x,-y identifies the systems for nn and −n-n.

For existence as a smooth action, the principal GG bundle S(ν2n)→Sb4S(\nu_{2n})\to S^4_b has clutching class 2n mod 2=02n\bmod2=0. It can therefore be glued equivariantly to G×D3G\times D^3. Van Kampen gives simple connectivity, and the local models give exactly the strata in (b). Automorphisms of the free boundary are absorbed using the free-boundary argument following Lemma 2.1, together with extension of a sphere diffeomorphism over D3D^3. This proves smooth realizability, with no assertion about strictly positive metrics. □

The cases n=0n=0 and n=1n=1 are S2×S4S^2\times S^4 and the twistor space CP3\mathbb{C}P^3, respectively. Proposition 4.2 proves the sphere-bundle description proposed in Liu’s Remark 4.5. It does not show that positive curvature forces ∣n∣≤1|n|\leq1.

Uniform boundary with one singular interior orbit

The interior tube in (c) is U=U1,1GU=U^G_{1,1} by Lemma 4.1. The region between it and the outer boundary tube is a principal bundle over S2×[0,1]S^2\times[0,1], hence an equivariant product in the interval direction. It can be absorbed into a collar. Consequently the entire manifold is

Mk=U1,1G∪φD(η⊞Lk).(25)M_k=U^G_{1,1}\cup_{\varphi}D(\eta\boxplus L^k). \tag*{(25)}

Proposition 4.3. The gluing (25) exists exactly when kk is odd. For each such kk it gives a closed simply connected smooth effective GG-manifold with the complete stratification (c). The equivariant attaching map introduces no further parameter.

Proof. The boundary Wk=S(η⊠Lk)W_k=S(\eta\boxtimes L^k) is a free GG-space over Sb2S_b^2. Local trivializations of LkL^k show that its principal GG-bundle clutching class is kk mod 2: the transition loop is the weight-kk rotation loop in T⊂GT\subset G. On the other side,

∂U1,1G=G×TS3⟶S3/T=S2\partial U_{1,1}^{G}=G\times_T S^3\longrightarrow S^3/T=S^2

is the extension of the Hopf circle bundle along T→GT\to G. Its clutching class is the nonzero element of π1(G)=Z/2\pi_1(G)=\mathbb{Z}/2. Therefore an equivariant boundary identification exists precisely for odd kk.

For odd kk, the Euler class 2a+kb2a+kb is primitive and WkW_k is simply connected. Both disk-bundle pieces in (25) are simply connected, so van Kampen gives π1(Mk)=0\pi_1(M_k)=0. The principal isotropy is trivial, and the stated slices contain no further exceptional strata. To compare two attaching maps, compose one with the inverse of the other on ∂U1,1G\partial U_{1,1}^{G}. Lemma 2.1 absorbs this boundary automorphism into that tube. This proves uniqueness for the fixed normal-bundle data. □\square

Theorem 4.4 (Integral classification of the odd family). For odd kk, the homology of MkM_k is torsion-free, b2=b4=2b_2=b_4=2, and Hodd(Mk;Z)=0H^{\mathrm{odd}}(M_k;\mathbb{Z})=0. There is a basis x,yx,y of H2(Mk;Z)H^2(M_k;\mathbb{Z}) restricting to a,ba,b on FF such that

∫Mkx3=k3−k4,∫Mkx2y=1−k22,∫Mkxy2=k,∫Mky3=−2.(26)\begin{aligned} \int_{M_k}x^3&=\frac{k^3-k}{4}, & \int_{M_k}x^2y&=\frac{1-k^2}{2}, & \int_{M_k}xy^2&=k, & \int_{M_k}y^3&=-2. \tag*{(26)} \end{aligned}
w2(Mk)=yˉ,PD⁡[F]=2x+ky.(27)w_2(M_k)=\bar y,\qquad\operatorname{PD}[F]=2x+ky. \tag*{(27)}
∫Mkp1(Mk)x=3k,∫Mkp1(Mk)y=−2.(28)\int_{M_k}p_1(M_k)x=3k,\qquad\int_{M_k}p_1(M_k)y=-2. \tag*{(28)}

Moreover Mk≅MℓM_k\cong M_\ell if and only if ∣k∣=∣ℓ∣|k|=|\ell|, and every MkM_k is rationally elliptic.

Proof. The boundary is S2×S3S^2\times S^3. In Mayer–Vietoris, the map H2(U)→H2(Wk)H^2(U)\to H^2(W_k) is an isomorphism, so restriction H2(Mk)→H2(F)H^2(M_k)\to H^2(F) is an isomorphism. The remaining terms give H3(Mk)=0H^3(M_k)=0 and a short exact sequence

0⟶Z⟶H4(Mk)⟶Z⟶0.0\longrightarrow\mathbb{Z}\longrightarrow H^4(M_k)\longrightarrow\mathbb{Z}\longrightarrow0.

Poincaré duality completes the additive computation. The restriction of TMkTM_k to FF and the normal Euler class give w2=yˉw_2=\bar y and PD⁡[F]=2x+ky\operatorname{PD}[F]=2x+ky.

The full cubic requires more than restriction to FF, since H4(Mk)→H4(F)H^4(M_k)\to H^4(F) is not injective. We compute it by the localization theorem [3, 4]. Let uu generate H2(BT;Q)H^2(BT;\mathbb{Q}) for the intrinsic rotation circle in SO⁡(3)\operatorname{SO}(3). The fixed set consists of two copies F+,F−F_+,F_- of Sb2S_b^2 over the boundary and two points q+,q−q_+,q_- on the interior orbit. For the canonical rational GG-equivariant lifts, the restrictions are

F+F_+F−F_-q+q_+q−q_-
xxu/2u/2−u/2-u/2ku/2ku/2−ku/2-ku/2
yybbbb−u-uuu
Euler classu(u+kb)u(u+kb)u(u−kb)u(u-kb)u3u^3−u3-u^3

Table 7.

The point restrictions follow from the boundary relation 2a+kb=02a+kb=0 in H2(Wk)H^2(W_k) and the fact that the two Weyl-related points have opposite restrictions. The point Euler sign is fixed by integrating the degree-four class u2yu^2y: its total integral is zero. Thus these signs do not require an arbitrary extra orientation choice.

For clarity, the contributions from the two surfaces and from the two points, respectively, are

ClassSurfacesPoints
x3x^3−k/4-k/4k3/4k^3/4
x2yx^2y1/21/2−k2/2-k^2/2
xy2xy^200kk
y3y^300−2-2

Table 8.

Their sums give (4.5). At F±F_{\pm} the equivariant Pontryagin class is 2u2±2kub2u^2 \pm2kub, and at q±q_{\pm} it is 3u23u^2. The surface contributions to p1x,p1yp_1x,p_1y are 0,40,4, while the point contributions are 3k,−63k,-6. This proves (4.7). The formulas also verify directly

∫M(2x+ky)x2=0,∫M(2x+ky)xy=1,∫M(2x+ky)y2=0,\int_M (2x+ky)x^2=0,\qquad\int_M (2x+ky)xy=1,\qquad\int_M (2x+ky)y^2=0,

which are the restrictions to FF.

The discriminant of the binary cubic fM(sx+ty)3f_M(sx+ty)^3 is

Δk=−27(k2−1).(29)\Delta_k=-27(k^2-1). \tag*{(29)}

It is unchanged by an integral change of basis or reversal of the ambient orientation. Hence a diffeomorphism forces ∣k∣=∣ℓ∣|k|=|\ell|. Conversely, reversing the orientation of Sb2S_b^2 identifies the normal data with kk replaced by −k-k; the corresponding boundary change is absorbed by Lemma 2.1. This gives Mk≅M−kM_k\cong M_{-k}. Equivalently, the full systems (4.5)–(4.7) can be compared by Wall–Jupp. Finally (4.4) is a double disk-bundle decomposition with b2=2b_2=2, so [8 Theorem A] proves rational ellipticity. □

For ∣k∣=1|k|=1, the cubic has a double root and MkM_k has the real homotopy type of S2×CP2S^2\times\mathbb{C}P^2. For ∣k∣>1|k|>1, its discriminant is negative and its real homotopy type is CP3#CP3\mathbb{C}P^3\#\mathbb{C}P^3. The nonzero class w2=yˉw_2=\bar y excludes every MkM_k from the spin models S6S^6, CP3\mathbb{C}P^3, S2×S4S^2\times S^4, FWF_W, and FEF_E. This is not an obstruction to positive curvature by itself.

An explicit representative is

M1≅P(L2⊕1⊕1)⟶S2.(30)M_1\cong\mathbb{P}(L^2\oplus1\oplus1)\longrightarrow S^2. \tag*{(30)}

It is induced from the circle action rotating the first homogeneous coordinate of CP2\mathbb{C}P^2. The fixed projective line gives the uniform boundary and the remaining coordinate fixed point gives the interior orbit. In a suitable basis its ring is

Z[x,y]/(x2,y3+2xy2),∫xy2=1.\mathbb{Z}[x,y]/(x^2,y^3+2xy^2),\qquad\int xy^2=1.

The quotient of a bi-invariant metric on GG and a Fubini–Study metric on CP2\mathbb{C}P^2 gives invariant nonnegative curvature. It does not give strict positivity. Thus Figure 2(c) is smoothly realizable, contrary to an interpretation of the diagram as topologically impossible; its positive-curvature realizability remains a separate question.

Doubling the quotient excludes two fixed points and an interior orbit.

Lemma 4.5. At an isolated GG-fixed point with tangent representation R3⊕R3\mathbb{R}^3\oplus\mathbb{R}^3, the quotient link is the round hemisphere S+2(1/2)S^2_+(1/2). At an interior orbit with slice (1,1)(1,1) it is S2(1/2)S^2(1/2).

Proof. Every invariant inner product on R3⊕R3\mathbb{R}^3\oplus\mathbb{R}^3 is carried to the standard inner product by an equivariant linear isometry. For the standard one write a representative of S5/SO(3)S^5/SO(3) as

z=eiθ(cos⁡t e1+isin⁡t e2),0≤t≤π/4,θ∈R/πZ.z=e^{i\theta}(\cos t\,e_1+i\sin t\,e_2),\qquad0\le t\le\pi/4,\quad\theta\in\mathbb{R}/\pi\mathbb{Z}.

Projecting the coordinate derivatives perpendicular to the orbits gives

dt2+cos⁡2(2t) dθ2=14(dρ2+sin⁡2ρ dϕ2),ρ=π/2−2t,ϕ=2θ.dt^2+\cos^2(2t)\,d\theta^2=\frac{1}{4}(d\rho^2+\sin^2\rho\,d\phi^2),\qquad\rho=\pi/2-2t,\quad\phi=2\theta.

The dependent-vector locus is the equator. The second assertion is the ordinary unit Hopf quotient; an invariant Hermitian form on the weight-one C2\mathbb{C}^{2} slice is likewise equivalent to the standard form. □

Proposition 4.6. Let ff count the boundary fixed points of the preceding type and let rr count interior singular points. Then f+2r≤3f+2r\le3. In particular case (e) cannot occur under (1.1).

Proof. Compactness gives sec⁡g≥κ>0\operatorname{sec}g\ge\kappa>0. Double XX along its whole boundary. Petrunin’s theorem [45] gives curvature at least κ\kappa on DXDX. Each boundary fixed point contributes one point whose link is the double of S ⁣+(1/2)S^{+}_{\!}(1/2), namely S2(1/2)S^{2}(1/2). Each interior singular point contributes two points. Their weighted-Hopf links have three-extent at most π/3\pi/3, as recalled in § 2. The four-point extent inequality therefore gives f+2r≤3f+2r\le3. In (e), f=2f=2 and r=1r=1, a contradiction. □

This proof doubles the metric quotient, not the six-manifold across its codimension-two singular set. It also shows that in the finite-fixed-set ball case the existence of two boundary fixed points automatically excludes an interior singular orbit. Lemma 4.1 then excludes exceptional orbits, and Liu’s Theorem 4.4 applies. Its equivariant homeomorphism to the linear six-sphere, together with Θ6=0\Theta_6=0, gives the ordinary diffeomorphism assertion for (d).

The excluded diagram is nevertheless smoothly realizable. Consider

Q3={[z1:⋯:z5]∈CP4:z12+⋯+z52=0},(31)Q^{3}=\{[z_{1}:\cdots:z_{5}]\in\mathbb{CP}^{4}:z_{1}^{2}+\cdots+z_{5}^{2}=0\}, \tag*{(31)}

with SO(3)\mathrm{SO}(3) rotating the first three coordinates. This quadric is the oriented Grassmannian SO(5)/(SO(2)×SO(3))\mathrm{SO}(5)/(\mathrm{SO}(2)\times\mathrm{SO}(3)) and is simply connected by the homogeneous-space homotopy sequence. Its fixed points are [0:0:0:1:i][0:0:0:1:i] and [0:0:0:1:−i][0:0:0:1:-i]. At z4=z5=0z_{4}=z_{5}=0 the conic is a single circle-isotropy orbit with normal weights (1,1)(1,1). When (z4,z5)≠0(z_{4},z_{5})\ne0, projective isotropy fixes the first three coordinates as a complex vector and is trivial or a circle; there are no exceptional orbits.

To identify the quotient, normalize a vector as (v,w)∈C3⊕C2(v,w)\in\mathbb{C}^{3}\oplus\mathbb{C}^{2}. The Gram matrix of the real and imaginary parts of vv determines its SO(3)\mathrm{SO}(3) orbit. The conditions

∥v∥2=1−∥w∥2,v⋅v=−w⋅w\lVert v\rVert^{2}=1-\lVert w\rVert^{2},\qquad v\cdot v=-w\cdot w

are realizable precisely when ∥w∥2+∣w⋅w∣≤1\lVert w\rVert^{2}+|w\cdot w|\le1. This is an S1S^{1}-invariant star-shaped four-ball in C2\mathbb{C}^{2}. Radial rescaling and the scalar circle quotient identify it with D4/S1≅B3D^{4}/S^{1}\cong B^{3}. The origin is the interior conic orbit and the boundary has exactly the two fixed points above. Thus Proposition 4.6 is an obstruction to positive curvature invariant under this action on Q3Q^{3}, not an assertion that the smooth diagram is impossible.

The remaining exceptional ball configuration is CP3\mathbb{CP}^{3}

Let II be the closed C2C_{2} interval in (f), including its boundary normalizer endpoint and its interior circle endpoint. Put Q=π−1(I)Q=\pi^{-1}(I).

Lemma 4.7. The inverse image QQ is a smooth embedded four-manifold equivariantly diffeomorphic to CP2\mathbb{CP}^{2} with its real SO(3)\mathrm{SO}(3) action. The interior singular sphere is its complex conic. The oriented normal line has Euler class ±h\pm h, where hh generates H2(CP2;Z)H^{2}(\mathbb{CP}^{2};\mathbb{Z}).

Proof. At the interior endpoint, QQ is the weight-two axis plane in the slice (1,2)(1,2). At the boundary endpoint, it is the standard plane in (4.1). It therefore closes smoothly. Align the principal subgroup as H=C2H=C_{2} about the zz axis. At the interior end the singular group is SO(2)z\mathrm{SO}(2)_{z}. At the boundary end it is O(2)x\mathrm{O}(2)_{x}, with x⊥zx\perp z: the generator of HH lies in the nonidentity component of this normalizer. This relative position, which is not expressed by the abstract symbols C2⊂O(2)C_{2}\subset\mathrm{O}(2) alone, is forced by (4.1).

The resulting cohomogeneity-one diagram is the real SO(3)\mathrm{SO}(3) action on CP2\mathbb{CP}^{2}, with singular orbits RP2\mathbb{RP}^{2} and the complex conic. The standard cohomogeneity-one classification [43, 5] gives the equivariant identification. The conic represents twice the projective-line class. Along it the ambient normal plane has circle weight one, and hence Euler number ±2\pm2. If e(νQ)=dhe(\nu Q)=dh, restriction to the conic gives 2d=±22d=\pm2, so d=±1d=\pm1. □

Lemma 4.8 (The complementary cap). The complement of an open normal disk bundle of QQ is equivariantly diffeomorphic to D(R3⊕R3)D(\mathbb{R}^{3}\oplus\mathbb{R}^{3}).

Proof. Choose the normal tube compatibly with the boundary-soul product description. Its quotient removes a neighborhood of the exceptional interval and leaves a ball with exactly one fixed point and a circle-isotropy side face. Remove a small linear disk about that fixed point. The remaining quotient is a product D2×[0,1]D^{2}\times[0,1] with its side-face structure, and there are no other orbit types in it. The interval vector field lifts to a smooth invariant vector field on the corresponding GG-manifold: on a face chart R2⊕C1\mathbb{R}^{2}\oplus\mathbb{C}^{1} choose the lift tangent to the face and use the squared normal radius as quotient normal coordinate. The field has no normal component along that face. On principal charts use a connection, and patch these lifts with an invariant partition of unity. Its flow gives an equivariant product cobordism. Thus the outer boundary is obtained from the boundary of the small linear fixed-point disk only by a collar. The small disk has representation R3⊕R3\mathbb{R}^{3}\oplus\mathbb{R}^{3}, proving the assertion. This is the smooth equivariant collar argument, not an assumption that an arbitrary topological cap has a standard smoothing. □

It follows that

M≅D(O(1)→CP2)∪S5D6.M\cong D\bigl(O(1)\to\mathbb{C}P^{2}\bigr)\mathop{\cup}_{S^{5}}D^{6}.

The negative Euler choice gives the same unoriented conclusion. The standard attachment is CP3\mathbb{C}P^{3}, decomposed as a tubular neighborhood of a hyperplane and the complementary disk. Changing the boundary attachment can alter this ordinary manifold only by a homotopy six-sphere, which is trivial by [36]. Hence M≅CP3M\cong\mathbb{C}P^{3}.

The action

A[z1:z2:z3:z4]=[A(z1,z2,z3)t:z4],A∈SO(3),A[z_{1}:z_{2}:z_{3}:z_{4}]=[A(z_{1},z_{2},z_{3})^{\mathsf{t}}:z_{4}],\qquad A\in\mathrm{SO}(3),

is Liu’s Example 5.3(3) and realizes exactly these strata with the positively curved Fubini–Study metric. The new conclusion is the uniqueness of the underlying smooth manifold, not the existence of that example. This completes Theorem 1.2.

SU(2) actions with orbit space S3

Assume G=SU(2)G=\mathrm{SU}(2), X≅S3X\cong S^{3}, and (1.1). By Liu’s Proposition 4.9, the singular orbits are G/TG/T spheres, their number is 1≤r≤31\le r\le3, and their union is exactly M{±1}M^{\{\pm1\}}. They are therefore totally geodesic. Every exceptional isotropy group is cyclic of odd order. At a singular orbit, the normal representation has a pair of coprime nonzero odd weights (p,q)(p,q), because the central element acts as −id- \mathrm{id} on the whole normal slice. These are existing results from [40]; the additional curvature restriction is the following theorem.

Adjoint-weight matching.

Theorem 5.1. Suppose an open exceptional edge has isotropy CkC_{k}, k>1k>1 odd, and transverse slice R⊕Cs\mathbb{R}\oplus C_{s}. Then

s≡±2(modk).(32)s\equiv\pm2\pmod{k}. \tag*{(32)}

Consequently, an exceptional edge incident to a one-valent singular vertex has order three.

Proof. A component of MCkM^{C_{k}} over the edge is a totally geodesic surface. Its residual rotation-circle action has a horizontal meridian connecting the two singular endpoints. Choose such a unit-speed geodesic γ:[0,L]→M\gamma:[0,L]\to M, perpendicular to the GG orbits. Its interior lies over the open edge. The endpoint isotropy circles can be aligned with the unique maximal torus containing CkC_{k}. Let m=su(2)⊖t\mathfrak{m}=\mathfrak{su}(2)\ominus\mathfrak{t}. For A∈mA\in\mathfrak{m}, write JA=A∗∣γJ_A=A^*|_\gamma for the Killing Jacobi field. Its span PtP_t is the adjoint weight-two plane. At an interior point, the CkC_k representation on the tangent space is R2⊕C2⊕Cs\mathbb{R}^2\oplus\mathbb{C}_2\oplus\mathbb{C}_s.

Suppose s≢±2(modk)s\not\equiv\pm2\pmod{k}. The weight-two isotypic component is then unique, so parallel transport along the pointwise CkC_k-fixed geodesic preserves PtP_t. Let Q0Q_0 be a fixed invariant inner product on m\mathfrak{m}. Since kk is odd and greater than one, the invariant metric on this real rotation representation is a scalar multiple of Q0Q_0. Therefore

⟨JA,JB⟩=f(t)2Q0(A,B),f(t)>0.(33)\langle J_A,J_B\rangle=f(t)^2Q_0(A,B),\qquad f(t)>0. \tag*{(33)}

In the identifications A↦JAA\mapsto J_A, the derivative operator on this plane is a(t)id⁡+b(t)Ja(t)\operatorname{id}+b(t)J, where JJ is its complex structure. The Wronskian of orbit Jacobi fields vanishes: the Killing identities express it as a multiple of ⟨γ′,[A∗,B∗]⟩\langle\gamma', [A^*,B^*]\rangle, which is zero because the bracket is orbit tangent and γ′\gamma' is horizontal. Thus b=0b=0. Differentiating (33) gives a=f′/fa=f'/f, whence

DtJA=(f′/f)JA,Dt2JA=(f′′/f)JA.D_tJ_A=(f'/f)J_A,\qquad D_t^2J_A=(f''/f)J_A.

The Jacobi equation now gives, for every nonzero AA,

K(γ′,JA)=−f′′/f>0.(34)K(\gamma',J_A)=-f''/f>0. \tag*{(34)}

At each endpoint the fields JAJ_A span the noncollapsed tangent plane of the singular sphere, so ff extends positively and smoothly to the endpoints. That sphere is totally geodesic. The first-variation formula for its orbit metric in the normal direction γ′\gamma' yields

f′(0)=f′(L)=0.f'(0)=f'(L)=0.

This contradicts the strict concavity in (34). The assumption of inequivalent characters is therefore impossible, proving (5.1).

At a one-valent vertex on an order-kk edge the slice weights are (1,k)(1,k); its transverse character is one. Thus 1≡±2(modk)1\equiv\pm2\pmod{k}. With k>1k>1 odd, this forces k=3k=3. □

The uniqueness of the weight-two isotypic summand is essential. When s≡±2s\equiv\pm2, the two rotation planes can mix under parallel transport, so the preceding scalar concavity argument does not apply. This is precisely the alternative retained by the theorem.

All remaining weights

There is no finite branching vertex, since all exceptional groups are cyclic. Every singular vertex has valency at most two. A one-vertex loop would force equal nontrivial weights and is excluded by coprimality. The remaining abstract possibilities are consequently the no-edge cases, an interval, a path, a bigon, a triangle, and an isolated vertex together with an interval or a bigon, as in the specialization of [37 Proposition 5.1].

For three singular vertices, a two-edge path has order-three leaves by Theorem 5.1, forcing middle weights (3,3)(3,3), which is impossible. In a genuine triangle order the labels a<b<ca<b<c. Along the cc edge, slice compatibility gives a≡±b(modc)a\equiv\pm b\pmod{c}. Hence c=a+bc=a+b, contradicting oddness. A genuine bigon together with an isolated singular vertex is excluded by the double branched cover along its cycle: [37 Proposition 2.14] preserves positive Alexandrov curvature, and its small-point conclusion gives the two vertices on the cycle and two lifts of the isolated vertex. Four such points contradict the extent bound. Thus the only three-vertex patterns are three isolated vertices or one interval together with an isolated vertex. In the second case its order is three.

For two singular vertices, the unordered slice pair (p,q)(p,q) is the same at both endpoints, each order greater than one being recorded by its edge. If p=qp=q, coprimality gives (1,1)(1,1). Otherwise take 1≤p<q1\leq p<q. The order-qq edge is genuine and has transverse character pp. By (5.1),

p≡±2(modq).p\equiv\pm2\pmod{q}.
Figure 4 patternRemaining labelsSmooth action
Two vertices(1,1)(1,1) or (2j−1,2j+1)(2j-1,2j+1), j≥1j \ge1N1,1GN^{G}_{1,1} or N2j−1,2j+1GN^{G}_{2j-1,2j+1} on S2×S4S^2 \times S^4.
Bigon and isolated vertex(1,1)(1,1) or (1,3)(1,3)Block action on FWF_W or commuting right block action on FEF_E.
Three-sided cycle(1,1,1)(1,1,1) or (1,1,3)(1,1,3)The same two flag actions after deleting labels one.

Table 5. The label conventions of Liu’s Figure 4, interpreted with order-one edges erased. The two-vertex row is a necessary positive-curvature candidate list; both three-vertex rows have the stated positively curved realizations.

Since pp is odd and 0<p<q0 < p < q, p=q−2p = q - 2. Thus the necessary list is

(p,q)=(1,1)or(p,q)=(2j−1,2j+1),j≥1.(35)(p,q) = (1,1) \quad\text{or} \quad(p,q) = (2j-1,2j+1), \qquad j \ge1. \tag*{(35)}

These pairs satisfy the matching congruences on both genuine edges, but that converse arithmetic statement is not a converse curvature theorem.

Two singular orbits: reconstruction of the weighted product actions

For a two-vertex bigon, the cycle is unknotted by the argument for graph 8 in [37 Section 6]. The interval and no-edge cases need no knot obstruction. Removing the two singular tubes therefore gives the orbifold product S2(p,q)×[0,1]S^2(p,q) \times[0,1]. Lifting the interval direction and applying Lemma 2.1 gives

M≅GUp,qG∪Up,qG=G×TS(R⊕Cp⊕Cq)=Np,qG.(36)M \cong_G U^G_{p,q} \cup U^G_{p,q} = G \mathbin{\times_T} S(\mathbb{R} \oplus\mathbb{C}_p \oplus\mathbb{C}_q) = N^G_{p,q}. \tag*{(36)}

For odd p,qp,q, the oriented five-plane bundle over G/T=S2G/T = S^2 has w2=p+q mod 2=0w_2 = p+q \bmod2 = 0, hence is trivial. The underlying manifold is S2×S4S^2 \times S^4.

Conversely, every positive coprime odd pair gives a smooth effective induced action with two singular orbits and the indicated slice pair. The diagonal quotient of a bi-invariant metric on GG and a round metric on the four-sphere gives invariant nonnegative curvature. In particular all pairs in (5.4) are smoothly realizable. This construction does not resolve invariant positive curvature, including for N1,1GN^G_{1,1}, which is the diagonal action in Liu’s Remark 4.12.

Three singular orbits with one exceptional interval

Proposition 5.2. A SU(2) action satisfying (1.1) with three singular orbits and an exceptional interval is equivariantly diffeomorphic to the commuting right block action on the Eschenburg flag FEF_E.

Proof. The interval has order three. Its inverse image is

A3=G×TS(R⊕C3)=P(1⊕L3)⟶S2,A_3 = G \mathbin{\times_T} S(\mathbb{R} \oplus\mathbb{C}_3) = \mathbb{P}(1 \oplus L^3) \longrightarrow S^2,

and the third singular orbit has slice (1,1)(1,1). The absolute normal characters at the two sections of A3A_3 are both one. Orient the normal line so that the first character is u=1u=1. Compatibility in Lemma 2.2 forces the other character to be v=1v=1, because −1≠1(mod3)-1 \ne1 \pmod3. Consequently its Euler class is one on a section and zero on a fiber. The line is equivariantly

νA3≅ρ∗L,(37)\nu A_3 \cong\rho^*L, \tag*{(37)}

where ρ:A3→S2\rho:A_3 \to S^2 is the bundle projection. In induced coordinates its disk bundle is

G×T(S(R⊕C3)×D(C1)).G \mathbin{\times_T} \left(S(\mathbb{R} \oplus\mathbb{C}_3) \times D(\mathbb{C}_1)\right).

The unit normal coordinate has weight one, so fixing its phase identifies the boundary equivariantly with G×S2G \times S^2.

Remove this edge neighborhood and a tube about the remaining singular orbit. The quotient complement is a free S2×[0,1]S^2 \times[0,1] product, and the principal SU(2) bundle on it is trivial. Absorb it into a collar. The complete manifold is therefore the union of the disk bundle in (37) and U1,1TGU^{T G}_{1,1}. Lemma 2.1 absorbs every attaching map on the free boundary. The endpoint characters determine the normal line, so there is a unique equivariant diffeomorphism type with these data. The explicit action verified below realizes this type on FEF_E.

This is a smooth reconstruction, not merely the agreement of Betti numbers recorded in Liu’s Example 5.5. Replacing 3 by any odd k≥3k \ge3 in the same disk-bundle construction still gives a smooth action with one CkC_k interval. Theorem 5.1 excludes its positive-curvature realization for k>3k > 3. Thus the restriction to order three is geometric, not a local smooth-slice restriction.

The Eschenburg model and its stabilizers

Use the determinant-preserving free torus action

(z,w)⋅g=diag⁡(z,w,zw)gdiag⁡(1,1,z2w2)−1,g∈SU⁡(3).(38)(z,w)\cdot g=\operatorname{diag}(z,w,zw)g\operatorname{diag}(1,1,z^{2}w^{2})^{-1},\qquad g\in\operatorname{SU}(3). \tag*{(38)}

Its quotient is FEF_E. The two diagonal factors have the same determinant, so (38) preserves SU⁡(3)\operatorname{SU}(3) even though neither factor separately need belong to it. The right upper-left SU⁡(2)\operatorname{SU}(2) block commutes with this torus action. To express the block action as a left action, write

A⋅[g]=[gdiag⁡(A,1)−1].A\cdot[g]=[g\operatorname{diag}(A,1)^{-1}].

For completeness, the stabilizer computation needed to identify the reconstruction is short. If a torus element fixes gg, the multisets {z,w,zw}\{z,w,zw\} and {1,1,(zw)2}\{1,1,(zw)^2\} agree. Two of z,w,zwz,w,zw must then be one, forcing z=w=1z=w=1. Thus the torus action is free. If v=(v1,v2,v3)v=(v_1,v_2,v_3) is the last column of gg, an element of the torus contributes to the block stabilizer precisely when

diag⁡(z,w,zw)v=(zw)2v.\operatorname{diag}(z,w,zw)v=(zw)^2v.

For the nonzero coordinates the respective conditions are

v1≠0: zw2=1,v2≠0: z2w=1,v3≠0: zw=1.v_1\ne0:\ zw^2=1,\qquad v_2\ne0:\ z^2w=1,\qquad v_3\ne0:\ zw=1.

All three nonzero coordinates give trivial stabilizer. A single nonzero coordinate gives a circle, accounting for three singular orbits. Supports {1,3}\{1,3\} and {2,3}\{2,3\} also give trivial stabilizer. On support {1,2}\{1,2\}, the equations yield z=wz=w and z3=1z^3=1, giving the noncentral order-three subgroup in the block. The parameter ∣v1∣2|v_1|^2 runs along its open quotient interval. The third circle orbit is isolated from that interval. This is exactly the pattern in Proposition 5.2.

Positive curvature for this model and invariance under the commuting block action are established in the Eschenburg construction and its symmetry analysis [9, 11, 22, 40]. We use that metric theorem directly, without reproducing its curvature computation. The ordinary flag FWF_W with the left upper-left block action likewise has three circle orbits and no exceptions and carries its invariant Wallach metric [51, 40].

The integral data for these flag manifolds are standard and can be obtained from biquotient transgression and the tangent representation; see [10, 16, 15]. In a common convenient convention they are

H∗(FW;Z)=Z[x,y]/(x3,y2−xy+x2),p1(FW)=0,(39)H^{*}(F_W;\mathbb{Z})=\mathbb{Z}[x,y]/(x^3,y^2-xy+x^2),\qquad p_1(F_W)=0, \tag*{(39)}
H∗(FE;Z)=Z[x,y]/(x3,y2−xy−x2),p1(FE)=8x2.(40)H^{*}(F_E;\mathbb{Z})=\mathbb{Z}[x,y]/(x^3,y^2-xy-x^2),\qquad p_1(F_E)=8x^2. \tag*{(40)}

Both have torsion-free cohomology, are spin, and are oriented by ∫x2y=1\int x^2y=1. To relate (40) to the torus characters a,ba,b in (38), take x=a+bx=a+b, y=by=b: the familiar relations are a2+3ab+b2=0a^2+3ab+b^2=0 and ab(a+b)=0ab(a+b)=0. For (39), the usual relations are a2+ab+b2=0a^2+ab+b^2=0, ab(a+b)=0ab(a+b)=0, and one may take x=ax=a, y=−by=-b. These changes of generators reconcile the two common sign conventions; they do not change the characteristic classes.

In particular the cubic polynomials in this convention are

∫FW(sx+ty)3=3s2t+3st2,∫FE(sx+ty)3=3s2t+3st2+2t3.\int_{F_W}(sx+ty)^3=3s^2t+3st^2,\qquad\int_{F_E}(sx+ty)^3=3s^2t+3st^2+2t^3.

Their real root types differ. More decisively for smooth recognition, the first Pontryagin class is zero on FWF_W and nonzero on FEF_E. Equal Betti numbers do not identify these manifolds.

Completion of the SU(2) classification

For three singular orbits and no exceptions, Liu’s corrected Theorem 4.11 already gives the equivariant action on FWF_W. For one singular orbit, the absence of a one-vertex loop implies there are no exceptions, so the same theorem gives the linear action

A(u,v)=(Au,Ad⁡Av),(u,v)∈S(C2⊕su(2)).(41)A(u,v)=(Au,\operatorname{Ad}_A v), \qquad(u,v)\in S(\mathbb{C}^{2}\oplus\mathfrak{su}(2)). \tag*{(41)}

There is no need to repeat Liu’s exceptional-free gluing proof. This representation has one singular sphere and no GG-fixed point. It should be distinguished from C2⊕Rtriv3\mathbb{C}^{2}\oplus\mathbb{R}^{3}_{\mathrm{triv}}, the triple suspension of the Hopf action, which has a fixed S2S^{2} and a ball quotient. Thus the explicit model in Liu’s Example 5.2(1), rather than a literal suspension description, is used in the spherical-quotient case.

The weight reductions, (36), and Proposition 5.2 prove Theorem 1.3. The remaining uncertainty in its two-orbit row is metric realizability on a determined smooth action, not a missing diffeomorphism classification.

Global consequences and the remaining reconstruction problem

Rational ellipticity throughout the symmetry class

Proof of Corollary 1.4. Liu’s orbit-space dichotomy gives X≅S3X\cong S^{3}, B3B^{3}, or B4B^{4}. If X≅S3X\cong S^{3}, rational ellipticity is Theorem A of Khalili Samani–Radeschi [37]. The B4B^{4} case and the SU(2) cases with boundary are identified by Liu with S6S^{6} or CP3\mathbb{C}P^{3} [40 Theorems 1.1–1.2 and 3.1].

For G=SO(3)G=\mathrm{SO}(3) and X≅B3X\cong B^{3}, § 4 exhausts the configurations. Cases (a), (d), and (f) give a sphere or CP3\mathbb{C}P^{3}, and case (e) is excluded. In case (b), the fibration S2→En→S4S^{2}\to E_n\to S^{4} implies rational ellipticity by the rational homotopy exact sequence. In case (c), the decomposition (25) is a double disk bundle with b2(Mk)=2b_{2}(M_k)=2. Theorem A of DeVito–Galaz-García–Kerin [8] applies: its six-dimensional hypothesis is b2≤3b_{2}\leq3, which is satisfied here. Thus all cases are rationally elliptic.

Liu gives χ(M)∈{2,4,6}\chi(M)\in\{2,4,6\}. A simply connected rationally elliptic space with positive Euler characteristic has no odd-degree rational cohomology [13]. Poincaré duality now gives χ(M)=2+2b2(M)\chi(M)=2+2b_{2}(M). Herrmann’s classification of rationally elliptic six-manifolds [30], restricted to this Betti range and positive Euler characteristic, yields the six real homotopy types stated in the corollary. □

The conclusion combines two independent geometric mechanisms: the boundaryless quotient theorem and the explicit disk-bundle reconstructions for quotients with boundary. No rational-ellipticity hypothesis is imposed on MM in addition to (1). On the other hand, rational ellipticity supplies no general torsion-freeness theorem. In the cyclic and ball cases, torsion-freeness follows from the actual decompositions; it is not inferred from Corollary 1.4.

What is recorded by a graph

The tables in § 3 record an abstract weighted graph. They distinguish singular vertices, branching exceptional vertices, and cyclic edge orders. They do not by themselves specify a smooth GG-manifold. Standard equivariant reconstruction [5 Chapters IV–V] requires the following additional information.

First one specifies a tame embedding of the graph into S3S^{3}, together with the actual isotropy subgroups, not only their abstract isomorphism types. At each vertex one specifies the slice representation and the inclusions of the edge groups into the vertex group. The normalizer transition data along an edge must intertwine these representations. These conditions include the transverse-character congruences of Lemma 3.2 and the normal Euler-class constraints of Lemma 2.2. Let BB be the closure of the complement of a small stratified neighborhood of the graph. Over BB one specifies a principal GG-bundle and its identifications with the boundary bundles of the local models. Principal SO⁡(3)\operatorname{SO}(3)-bundles over a three-dimensional CW complex are classified by w2∈H2(B;Z/2)w_2 \in H^2(B;\mathbb{Z}/2); principal SU⁡(2)\operatorname{SU}(2)-bundles over such a complex are trivial [32, 41]. Finally one specifies the equivariant attaching maps, modulo automorphisms extending over the pieces. In the cyclic configurations, the unknotting input and Lemma 2.1 remove these apparent extra choices. That reduction has not been established for every branching graph.

This distinction also explains why the present graphs should not be confused with GKM graphs. Under the hypotheses of GKM theory, a higher-rank torus action can encode integral equivariant characteristic classes in a labelled graph; the six-dimensional recognition theorem of Goertsches–Konstantis–Zoller is a particularly relevant example [15]. An abstract finite-isotropy graph for a rank-one nonabelian action does not automatically satisfy those hypotheses or retain the same information. Its edge orders alone do not determine an integral degree-two lattice.

Integral completion of the branching cases

For a specified reconstruction, Mayer–Vietoris determines the integral homology from the inclusion maps of the boundary pieces. These maps, rather than just the ranks of the groups, decide torsion and the primitive lattice

H2(M;Z)↪⨁iH2(Si;Z).H^2(M;\mathbb{Z}) \hookrightarrow\bigoplus_i H^2(S_i;\mathbb{Z}).

The rational image is the hyperplane in (11). In a rank-one lollipop case, the integer tt in (18) measures information that this rational hyperplane does not detect. Once the lattice is known, Proposition 2.3 gives the cubic cup form and the Pontryagin covector. The transition functions of the tangent bundle, or an equivalent integral normal-bundle calculation, are still needed to determine w2w_2.

When homology is torsion-free, the Wall–Jupp theorem [50, 33] converts the resulting tuple (3) into an ordinary diffeomorphism classification. If torsion occurs, one must retain its additional linking and normal data and use the appropriate classification framework, rather than apply the torsion-free theorem [57]. In particular, rational cohomology of a sphere does not imply that a theta-graph manifold is a homotopy sphere. If torsion-freeness is separately established, simple connectivity and the Hurewicz theorem do give a homotopy sphere, and Θ6=0\Theta^6 = 0 identifies it with S6S^6 [36].

For comparison, the standard characteristic data of several recognition models are recorded below. The degree-two generators and the orientation are part of each entry. The flag data are given in (5.8)–(5.9); the data for EnE_n, Pa,bP_{a,b}, and MkM_k have already been computed.

ModelIntegral intersection dataCharacteristic classes
S6S^6H2=H4=0H^2 = H^4 = 0w2=p1=0w_2 = p_1 = 0
S2×S4S^2 \times S^4x2=0x^2 = 0, ⟨xy,[M]⟩=1\langle xy,[M]\rangle= 1w2=p1=0w_2 = p_1 = 0
CP3\mathbb{C}P^3⟨x3,[M]⟩=1\langle x^3,[M]\rangle= 1w2=0w_2 = 0, p1=4x2p_1 = 4x^2
S2×CP2S^2 \times\mathbb{C}P^2b2=a3=0b^2 = a^3 = 0, ⟨a2b,[M]⟩=1\langle a^2b,[M]\rangle= 1w2=aˉw_2 = \bar{a}, p1=3a2p_1 = 3a^2
CP3#CP3\mathbb{C}P^3\#\mathbb{C}P^3xy=0xy = 0, ⟨x3,[M]⟩=⟨y3,[M]⟩=1\langle x^3,[M]\rangle= \langle y^3,[M]\rangle= 1w2=0w_2 = 0, p1=4x2+4y2p_1 = 4x^2 + 4y^2

Table 9.

The product and projective-space formulas follow from the standard tangent-bundle identities [41]; connected-sum characteristic numbers are additive. These are recognition data, not assertions of positive-curvature existence for all entries.

Overlaps and limits of the classification

Different orbit stratifications can occur on the same smooth manifold. The family EnE_n appears in the uniform-boundary ball case, while Proposition 3.4 realizes E(k2−1)/8E_{(k^2-1)/8} with a dihedral lollipop for every odd k≥3k \ge3. Likewise, M1M_1 is the nontrivial unoriented linear CP2\mathbb{C}P^2-bundle over S2S^2; it is diffeomorphic to P1,2P_{1,2}, although the displayed actions have different orbit-space topology. The assertion follows from their clutching classes modulo three, or from their integral characteristic data. These overlaps do not identify the actions equivariantly.

A complete classification under (1.1) would require two further steps. In the explicitly reconstructed families, one must determine which parameters admit invariant positive curvature. In the remaining SO⁡(3)\operatorname{SO}(3) branching families, one must first finish integral reconstruction and then address curvature. Neither task follows merely from the finite graph catalogue or the bound b2≤2b_2 \le2.

The frequently suggested finite target

S6,CP3,S2×S4,FW,FE(42)S^6,\qquad\mathbb{C}P^3,\qquad S^2 \times S^4,\qquad F_W,\qquad F_E \tag*{(42)}

is therefore a classification target, not the conclusion of this paper. For a torsion-free candidate, agreement of the Wall–Jupp data with an entry in (6.1) is an exact smooth-recognition criterion. The missing geometric work is to force that agreement and to decide the remaining invariant metric questions. The next section addresses part of the latter issue without confusing a restricted curvature obstruction with a complete classification.

Curvature obstructions for the diagonal product actions

The preceding classifications leave several actions on S2×S4S^2 \times S^4 as smooth candidates. We now isolate what can be proved about the two diagonal actions discussed by Liu. The restrictions below do not require or imply a positive-curvature metric on any of the other smooth models.

The two actions and the space of invariant metrics

For the SO⁡(3)\operatorname{SO}(3) action, write

M=S2×S(R3⊕R2),A(x,y,z)=(Ax,Ay,z).(43)M = S^2 \times S(\mathbb{R}^3 \oplus\mathbb{R}^2),\qquad A(x,y,z) = (Ax,Ay,z). \tag*{(43)}

The orbit map and singular set are

q(x,y,z)=(x⋅y,z)∈B3,F={(x,tx,z):t2+∣z∣2=1}≅S2×S2.(44)q(x,y,z) = (x \cdot y,z) \in B^3,\qquad F = \{(x,tx,z):t^2 + |z|^2 = 1\} \cong S^2 \times S^2. \tag*{(44)}

Indeed, the Gram data ∣x∣=1|x| = 1, ∣y∣2=1−∣z∣2|y|^2 = 1 - |z|^2, and x⋅y=tx \cdot y = t classify pairs of vectors up to SO⁡(3)\operatorname{SO}(3). Isotropy is trivial when x,yx,y are independent and is SO⁡(2)\operatorname{SO}(2) when they are dependent. This is the n=0n = 0 member of case (b). Its normal bundle is νF≅pr⁡1∗TS2\nu F \cong\operatorname{pr}_1^*TS^2, with Euler class 2a2a.

For the SU⁡(2)=Sp⁡(1)\operatorname{SU}(2) = \operatorname{Sp}(1) action, quaternionic coordinates give

M=S(Im⁡H)×S(H⊕R),a(u,q,t)=(aua−1,aq,t).(45)M = S(\operatorname{Im}\mathbb{H}) \times S(\mathbb{H} \oplus\mathbb{R}),\qquad a(u,q,t) = (aua^{-1},aq,t). \tag*{(45)}

The two singular orbits are q=0q = 0, t=±1t = \pm1. They are exactly the fixed components of the center. For q≠0q \ne0, the quotient coordinates are

(qˉuq∣q∣2,t)∈S2×(−1,1).\left(\frac{\bar{q}uq}{|q|^2},t\right) \in S^2 \times(-1,1).

Collapsing each end sphere gives S3S^3. This is N1,1SU⁡(2)N^{\operatorname{SU}(2)}_{1,1}. The center acts nontrivially on the quaternionic coordinate, so this action does not descend to the SO⁡(3)\operatorname{SO}(3) action in (7.1).

Both actions have an induced description G×TS4G \times_T S^4. A connection on G→G/TG \to G/T identifies their tangent bundle with

G×T((S4×m)⊕TS4),m=g⊖t.G \times_T\left((S^4 \times\mathfrak{m}) \oplus TS^4\right),\qquad\mathfrak{m} = \mathfrak{g} \ominus\mathfrak{t}.

Consequently, an arbitrary invariant metric is exactly a smooth TT-equivariant positive-definite block form

gz((A,U),(B,V))=az(A,B)+cz(A,V)+cz(B,U)+bz(U,V),(46)g_z((A,U),(B,V)) = a_z(A,B) + c_z(A,V) + c_z(B,U) + b_z(U,V), \tag*{(46)}

where A,B∈mA,B \in\mathfrak{m} and U,V∈TzS4U,V \in T_zS^4. This is a parametrization of all invariant metrics, not a connection-metric ansatz. In particular the cross term czc_z cannot in general be discarded. At a pole of the SU⁡(2)\operatorname{SU}(2) slice, equivariance forces this cross term to vanish because the orbit tangent has weight two and the normal slice has weight one. It does not force the cross term to vanish away from the poles.

Nor does invariance force either product projection to be a Riemannian submersion with totally geodesic fibers. For example, (2+t)2gS2+gS4(2+t)^2g_{S^2}+g_{S^4} is invariant under (7.3) but has varying orbit radii. The invariant function ⟨qˉuq,iqˉ⟩\langle\bar{q}uq,i\bar{q}\rangle can be used in a positive conformal factor to break the full commuting right-Sp⁡(1)\operatorname{Sp}(1) symmetry. In (7.1), the positive invariant conformal factor 2+z12+z_1 breaks the standard suspension-coordinate O⁡(2)\operatorname{O}(2) symmetry. These examples are not positive-curvature constructions; they show that the additional hypotheses in the obstructions below are genuine restrictions.

Cheeger deformations and conformal changes

Proposition 7.1. For either action, let g0g_0 be a GG-invariant product metric on S2×S4S^2 \times S^4. Every diagonal Cheeger deformation gsg_s, s≥0s \ge0, has a zero-curvature two-plane at every point. No conformal rescaling of any gsg_s has everywhere positive sectional curvature.

Proof. Use the Cheeger submersion [56]

σ:(M×G,g0+s−1Q)⟶(M,gs),σ(p,a)=a−1p,\sigma:(M \times G,g_0+s^{-1}Q)\longrightarrow(M,g_s),\qquad\sigma(p,a)=a^{-1}p,

where QQ is bi-invariant and s>0s>0. At pp choose 0≠X∈TS20\ne X\in TS^2 and 0≠Y∈TS40\ne Y\in TS^4 orthogonal to the GG-orbit in the second factor. Such YY exists because those orbits have dimension at most three. Define L∈gL\in\mathfrak{g} by Q(L,Z)=g0(X,Z∗)Q(L,Z)=g_0(X,Z^*). At (p,1)(p,1) the vectors

X^=(X,−sL),Y^=(Y,0)\widehat{X}=(X,-sL),\qquad\widehat{Y}=(Y,0)

are horizontal. Their ambient curvature numerator is zero: the MM components form a mixed product plane and the second GG component vanishes.

The O’Neill term also vanishes. The metric-dual one-form ηZ\eta_Z of a vertical fundamental field is a sum of forms from MM and GG. Its MM part is the sum of one-forms pulled back from the two product factors. Hence dηZ(X^,Y^)=0d\eta_Z(\widehat{X},\widehat{Y})=0 for every ZZ. For horizontal extensions this implies [X^,Y^]vert=0[\widehat{X},\widehat{Y}]^{\mathrm{vert}}=0. O’Neill’s horizontal curvature formula [42] therefore gives a zero plane in (M,gs)(M,g_s). The case s=0s=0 is immediate.

More generally, suppose a metric gg on a compact manifold has a nonpositive plane at every point. At a minimum of a smooth function ff, its gradient vanishes and its Hessian is nonnegative. The conformal curvature formula [44] gives, for a gg-orthonormal basis U,VU,V of such a plane,

Ke2fg(U∧V)=e−2f(Kg(U∧V)−Hess⁡gf(U,U)−Hess⁡gf(V,V))≤0.K_{e^{2f}g}(U\wedge V)=e^{-2f}\bigl(K_g(U\wedge V)-\operatorname{Hess}_g f(U,U)-\operatorname{Hess}_g f(V,V)\bigr)\le0.

Apply this to each gsg_s. □

The same minimum argument excludes a warped product over a closed base with nontrivial spherical fiber: the mixed curvature is −f−1Hess⁡f-f^{-1}\operatorname{Hess} f at a minimum of the warping function. It does not apply to the full block form (7.4).

Both totally geodesic product fibrations are excluded.

Theorem 7.2. There is no positively curved metric on S2×S4S^2 \times S^4 for which either standard product projection is a Riemannian submersion with totally geodesic fibers. No invariance assumption is needed.

Proof. We use O’Neill’s mixed curvature formula for totally geodesic fibers [42]:

K(X,V)=∣AXV∣2(47)K(X,V)=\lvert A_XV\rvert^2 \tag*{(47)}

for horizontal and vertical unit vectors. For the projection onto S2S^2, the linear map AXA_X sends the four-dimensional vertical space into the one-dimensional horizontal space orthogonal to XX. It has a nonzero kernel, producing a zero mixed plane. Consider the projection onto S4S^4. We first show that the two-sphere fibers would necessarily be round; this is not an initial assumption. Totally geodesic fibers are mutually isometric under horizontal transport. The curvature of this transport connection is a vertical Killing field on each fiber, represented, up to the conventional sign and factor, by 2AXY2A_XY. At a point of a fiber, if the values of these Killing fields failed to span its tangent plane, a nonzero vertical VV would be perpendicular to all AXYA_XY. The identity ⟨AXV,Y⟩=−⟨V,AXY⟩\langle A_XV,Y\rangle=-\langle V,A_XY\rangle would imply AXV=0A_XV=0 for all XX, contradicting (7.5). The fiber isometry group therefore acts transitively. Its Gaussian curvature is constant, and Gauss–Bonnet makes that constant positive. All fibers are isometric round spheres of one radius.

The connection thus reduces to an SO⁡(3)\operatorname{SO}(3) connection on the round sphere bundle. Let Ω\Omega be its curvature, identified fiberwise with a two-form on the base valued in R3\mathbb{R}^3. At a unit sphere point pp, the horizontal O’Neill tensor is

AXY=−12Ω(X,Y)×p.A_XY=-\frac{1}{2}\Omega(X,Y)\times p.

For every nonzero w∈R3w\in\mathbb{R}^3, the two-form Ωw=⟨Ω,w⟩\Omega_w=\langle\Omega,w\rangle on the four-dimensional base must be nondegenerate. Otherwise choose X≠0X\ne0 in its kernel, p⊥wp\perp w, and the vertical vector V=w×pV=w\times p. Since p×V=wp\times V=w, the displayed formula implies AXV=0A_XV=0, again contradicting positivity. This is the fatness condition for the associated two-sphere bundle.

The standard coupling-form construction for a fat round two-sphere bundle now gives a symplectic form on its total space; see Fine–Panov [14 Section 2.2]. We use that theorem directly. In a trivialization its familiar local expression is the fiber area form minus d⟨p,γ⟩d\langle p,\gamma\rangle, with γ\gamma the connection one-form, and nondegeneracy is precisely the condition on Ωw\Omega_w above. But

H2(S2×S4;R)=Rα,α2=0.H^2(S^2\times S^4;\mathbb{R})=\mathbb{R}\alpha,\qquad\alpha^2=0.

Every degree-two class has zero cube, whereas a symplectic form on a closed six-manifold has nonzero cubic integral. This contradiction proves the theorem. □

The argument extracts fatness from the submersion hypothesis. The absence of a symplectic form by itself is not an obstruction to positive sectional curvature: S6S^6 is the simplest warning against that inference.

Additional symmetries forbidden by positive curvature.

Proposition 7.3. For the action (7.1), a positively curved invariant metric cannot extend to the diagonal O⁡(3)\operatorname{O}(3) action. Moreover no nonidentity standard transformation

(x,y,z)⟼(x,y,Rz),R∈O⁡(2),(x,y,z)\longmapsto(x,y,Rz),\qquad R\in\operatorname{O}(2),

can be an isometry of a positively curved metric on MM. For (7.3), a positively curved invariant metric cannot also be invariant under the full commuting right-Sp⁡(1)\operatorname{Sp}(1) action on qq.

Proof. Fixed components of isometries are totally geodesic. We apply Wilking’s connectedness theorem [52]: a closed codimension-kk totally geodesic submanifold of a positively curved nn-manifold has (n−2k+1)(n-2k+1)-connected inclusion.

A reflection in the diagonal O⁡(3)\operatorname{O}(3) action fixes S1×S3S^1\times S^3, of codimension two. Its inclusion would be three-connected, contradicting its nontrivial fundamental group. A nonidentity rotation of the suspension coordinates fixes S2×S2S^2\times S^2, also of codimension two. Its second homotopy group has rank two, whereas π2(M)\pi_2(M) has rank one. A reflection of the suspension coordinates fixes S2×S3S^2\times S^3, of codimension one. Five-connectedness would identify its π3≅Z2\pi_3\cong\mathbb{Z}^2 with π3(M)≅Z\pi_3(M)\cong\mathbb{Z}, again impossible.

For the quaternionic action, adjoining right-Sp⁡(1)\operatorname{Sp}(1) makes the action cohomogeneity one: the regular orbits are the level sets of tt. Verdiani’s classification of positively curved, even-dimensional, simply connected cohomogeneity-one manifolds [49], also discussed in [25], would make MM a compact rank-one symmetric space. The integral cohomology ring of S2×S4S^2 \times S^4 excludes this possibility. □

An action-specific boundary identity for the SO(3) action

The union FF of singular orbits in (7.2) is not the fixed set of a single subgroup. It need not be totally geodesic. The following calculation keeps exactly the second-fundamental-form term that this distinction permits.

Write F=S2×BF = S^2 \times B, B=S2B = S^2, with orbit tangent bundle V\mathcal{V} and base tangent bundle TBTB. The isotropy TT has weights one on V\mathcal{V} and νF\nu F, and weight zero on TBTB. Thus every invariant metric induces

g∣F=f(b)2gS2(1)+h,f:B→(0,∞),(48)g|_F = f(b)^2g_{S^2(1)} + h,\qquad f:B \to(0,\infty), \tag*{(48)}

where ff is the intrinsic singular-orbit radius. The second fundamental form satisfies

II(V,V)=0,II(TB,TB)=0.(49)\mathrm{II}(\mathcal{V},\mathcal{V}) = 0,\qquad\mathrm{II}(TB,TB) = 0. \tag*{(49)}

Indeed, the symmetric square of weight one contains only weights zero and two, not weight one; the zero-weight summand also cannot map equivariantly to the normal plane.

Along the section {e3}×B\{e_3\} \times B, choose unit orbit and normal frames; they exist globally because both plane bundles restrict trivially there. Using the complex structures defined by the circle action, the remaining block is encoded by

β∈Ω1(B;C),∣II(U,X)∣2=∣β(X)∣2(∣U∣=1, U∈V).(50)\beta\in\Omega^1(B;\mathbb{C}),\qquad|\mathrm{II}(U,X)|^2 = |\beta(X)|^2\qquad(|U|=1,\ U\in\mathcal{V}). \tag*{(50)}

Equivariance extends this identity to every orbit. A change of unit normal frame multiplies β\beta by a unit complex function, leaving its real rank and β∗β\beta^*\beta unchanged.

Proposition 7.4. For orthonormal orbit vectors U,VU,V and hh-orthonormal base vectors X,YX,Y, the ambient sectional curvatures along FF satisfy

Kg(U,V)=1−∣df∣h2f2,(51)K_g(U,V) = \frac{1-|df|_h^2}{f^2}, \tag*{(51)}
Kg(X,Y)=Kh,(52)K_g(X,Y) = K_h, \tag*{(52)}
Kg(U,X)=−Hess⁡hf(X,X)f+∣β(X)∣2.(53)K_g(U,X) = -\frac{\operatorname{Hess}_h f(X,X)}{f} + |\beta(X)|^2. \tag*{(53)}

These identities concern the indicated planes, not all planes in TM∣FTM|_F.

Proof. Apply the standard warped-product curvature formulas to (7.6), and then the Gauss equation [44]. The pure blocks of the second fundamental form vanish by (7.7). The remaining Gauss correction on a mixed plane is +∣II(U,X)∣2+|\mathrm{II}(U,X)|^2, which is (7.8). □

Corollary 7.5. If gg is invariant under (7.1) and sec⁡g>0\sec_g > 0, then

Kh>0,∣df∣h<1,S:=β∗β−f−1Hess⁡hf>0.(54)K_h > 0,\qquad|df|_h < 1,\qquad S := \beta^*\beta- f^{-1}\operatorname{Hess}_h f > 0. \tag*{(54)}

The form βb:TbB→C\beta_b:T_bB \to\mathbb{C} is a real-linear isomorphism at every minimum of ff, and at every degeneracy point it satisfies

Hess⁡hf(X,X)<0for all 0≠X∈ker⁡βb.\operatorname{Hess}_h f(X,X) < 0\qquad\text{for all }0\ne X\in\ker\beta_b.

In particular no such positive metric has constant singular-orbit radius, or has rank⁡Rβ≤1\operatorname{rank}_{\mathbb{R}}\beta\le1 everywhere. In addition,

∫B∣β∣h2 dAh>∫B∣dlog⁡f∣h2 dAh.(55)\int_B |\beta|_h^2\,dA_h > \int_B |d\log f|_h^2\,dA_h. \tag*{(55)}

If sec⁡g≥κ>0\sec_g \ge\kappa> 0, the right-hand side can be increased by 2κArea⁡(B,h)2\kappa\operatorname{Area}(B,h), with ≥\ge in place of >>.

Proof. The pointwise inequalities follow directly from Proposition 7.4. At a minimum the Hessian is nonnegative, so S>0S>0 forces injectivity of β\beta. A globally injective β:TS2→S2×C\beta:TS^2\to S^2\times\mathbb{C} would trivialize TS2TS^2, contradicting its Euler number two. Hence β\beta degenerates somewhere, ruling out constant ff. Tracing S>0S>0 and integrating gives (7.13), since

∫Bf−1Δhf dAh=∫Bf−2∣df∣h2 dAh.(56)\int_B f^{-1}\Delta_h f\,dA_h=\int_B f^{-2}|df|_h^2\,dA_h. \tag*{(56)}

The quantitative assertion follows by tracing the lower bound S≥κhS\geq\kappa h.

Even the unit round product metric has a nonzero mixed second fundamental form on FF. Its induced metric is (1+t2)gS2(1)+gB(1)(1+t^2)g_{S^2(1)}+g_B(1). At t=0t=0, x=e3x=e_3, take the orbit tangent U=(e1,0,0)U=(e_1,0,0) and the base tangent X=(0,e3,0)X=(0,e_3,0) corresponding to increasing tt. Differentiation of (7.2) gives II(U,X)=(0,e1,0)≠0\mathrm{II}(U,X)=(0,e_1,0)\ne0. In (7.11), the Hessian term −1-1 and the normal term +1+1 cancel. Discarding the latter would give an incorrect curvature calculation.

Compatibility of the boundary inequalities.

Proposition 7.6. There are smooth data (h,f,β)(h,f,\beta) on B=S2B=S^2 satisfying all the inequalities in (7.12). They occur as the induced metric and mixed second fundamental form along FF for a smooth invariant metric on MM. This assertion does not give positive curvature on all ambient planes.

Proof. Write b=(b1,b2,r)∈S2b=(b_1,b_2,r)\in S^2, ζ=b1+ib2\zeta=b_1+ib_2, and choose 0<ε≤120<\varepsilon\leq\frac{1}{2}. Set

h=gS2(1),f=1−εr,β=ε((1+r)dζ−ζ dr).(57)h=g_{S^2(1)},\qquad f=1-\varepsilon r,\qquad\beta=\sqrt{\varepsilon}\bigl((1+r)d\zeta-\zeta\,dr\bigr). \tag*{(57)}

These are globally smooth. Direct calculation gives

β∗β=ε(1+r)2h,Hess⁡hf=εrh,∣df∣h≤ε<1.\beta^*\beta=\varepsilon(1+r)^2h,\qquad\operatorname{Hess}_h f=\varepsilon r h,\qquad|df|_h\leq\varepsilon<1.

Thus

S=ε((1+r)2−r1−εr)h>0.S=\varepsilon\left((1+r)^2-\frac{r}{1-\varepsilon r}\right)h>0.

For r<0r<0 both contributions are nonnegative and their sum is positive. For 0≤r≤10\leq r\leq1, the expression in parentheses is at least (1+r)2−2r=1+r2(1+r)^2-2r=1+r^2. Also Kh=1K_h=1. The degeneracy of β\beta is at the maximum of ff, not at its minimum.

To realize these tensors, equip E=pr⁡1∗TS2→FE=\operatorname{pr}_1^*TS^2\to F with its standard fiber metric and an invariant connection. Prescribe gF=f2gS2+hg_F=f^2g_{S^2}+h and let II\mathrm{II} have the pure blocks (7.7) and, for an orbit coordinate vector u∈TxS2u\in T_xS^2, mixed block

II(u,X)=f(Re⁡β(X)u+Im⁡β(X)Jxu)∈E(x,b).\mathrm{II}(u,X)=f\bigl(\operatorname{Re}\beta(X)u+\operatorname{Im}\beta(X)J_xu\bigr)\in E_{(x,b)}.

In an equivariant tubular neighborhood, set for small n∈Epn\in E_p

gn(Ah,Bh)=gF(A,B)−2⟨II(A,B),n⟩,g_n(A^h,B^h)=g_F(A,B)-2\langle\mathrm{II}(A,B),n\rangle,

use the Euclidean metric in the fiber directions, and make horizontal and vertical directions orthogonal. Compactness gives a uniform neighborhood on which this tensor is positive definite. Its normal derivative at the zero section realizes the prescribed second fundamental form. An invariant cutoff and a convex interpolation with any invariant metric extend it to all of MM without changing it near FF.

The planes in Proposition 7.4 consequently have positive curvature along FF. Curvatures of planes containing normal vectors, general oblique planes, and planes in the regular part have not been controlled. This is exactly the limitation of the construction.

A first-order restriction without symmetry.

Proposition 7.7. Let gsg_s be a smooth family of metrics on S2×S4S^2 \times S^4 with g0g_0 the round product metric. If the first variation of sectional curvature is nonnegative on every initially mixed plane, it vanishes on every such plane. In particular this vanishing is necessary if sec⁡gs>0\operatorname{sec}_{g_s} > 0 for every sufficiently small s>0s > 0.

Proof. The product of a great circle in each sphere is a flat totally geodesic torus TT for g0g_0. Its second fundamental form for gsg_s is O(s)O(s), so its quadratic Gauss correction is O(s2)O(s^2). The first variation of its intrinsic Gaussian curvature therefore equals that of the ambient sectional curvature on TTT T. Differentiating Gauss–Bonnet gives

∫Tdds∣s=0Kgs(TpT) dAg0=0,\int_T \left.\frac{d}{ds}\right|_{s=0} K_{g_s}(T_pT)\,dA_{g_0}=0,

since the initial curvature is zero. A nonnegative integrand must vanish identically. Every mixed plane is tangent to a torus of this form. Positivity for s>0s > 0 implies nonnegativity of the right derivative at zero, proving the last assertion.

This rules out a strictly positive first-order mixed-curvature direction, not a higher-order perturbation. The distinction is relevant to the recent perturbative construction reported in [6], although its manifold and prescribed symmetry are different.

The unresolved metric questions.

For both diagonal actions, the quotient and the underlying smooth manifold are now explicit. The arguments above exclude product and warped-product constructions, diagonal Cheeger deformations and their conformal classes, both totally geodesic product fibrations, and certain natural enlarged symmetry groups. The SO(3)\mathrm{SO}(3) action has the additional singular-orbit-radius obstruction of Corollary 7.5.

None of these hypotheses is forced by the prescribed rank-one symmetry. The full metric space is (7.4), and Proposition 7.6 shows that even the boundary restrictions have nonconstant compatible data. Thus the existence of an arbitrary positively curved invariant metric for either action is not decided by this paper. The smooth candidate classifications in Sections 3–5 and the restricted metric obstructions in this section are logically separate conclusions.

References

Email address: [email protected]

References

  1. [1]A. Adem, J. Leida, and Y. Ruan, Orbifolds and Stringy Topology, Cambridge Tracts in Mathematics 171, Cambridge University Press, Cambridge, 2007.DOI
  2. [2]M. M. Alexandrino and R. G. Bettio, Lie Groups and Geometric Aspects of Isometric Actions, Springer, Cham, 2015.DOI
  3. [3]M. F. Atiyah and R. Bott, The moment map and equivariant cohomology, Topology 23 (1984), no. 1, 1–28. doi:10.1016/0040-9383(84)90021-1.DOI
  4. [4]N. Berline and M. Vergne, Classes caractéristiques équivariantes. Formule de localisation en cohomologie équivariante, C. R. Acad. Sci. Paris Sér. I Math. 295 (1982), 539–541.
  5. [5]G. E. Bredon, Introduction to Compact Transformation Groups, Pure and Applied Mathematics 46, Academic Press, New York–London, 1972.
  6. [6]S. Brendle and P.-K. Hung, A metric on S² × S² with positive sectional curvature, preprint, arXiv:2608.19068v2, revised August 24, 2026.
  7. [7]O. Dearricott, A 7-manifold with positive curvature, Duke Math. J. 158 (2011), no. 2, 307–346.DOI
  8. [8]J. DeVito, F. Galaz-García, and M. Kerin, Manifolds that admit a double disk-bundle decomposition, Indiana Univ. Math. J. 72 (2023), no. 4, 1503–1551. doi:10.1512/iumj.2023.72.9408; arXiv:2008.10278v2.DOI
  9. [9]J.-H. Eschenburg, New examples of manifolds with strictly positive curvature, Invent. Math. 66 (1982), no. 3, 469–480.DOI
  10. [10]J.-H. Eschenburg, Cohomology of biquotients, Manuscripta Math. 75 (1992), no. 2, 151–166.DOI
  11. [11]J.-H. Eschenburg, Inhomogeneous spaces of positive curvature, Differential Geom. Appl. 2 (1992), no. 2, 123–132.DOI
  12. [12]F. Fang and X. Rong, Homeomorphism classification of positively curved manifolds with almost maximal symmetry rank, Math. Ann. 332 (2005), 81–101.
  13. [13]Y. Félix, S. Halperin, and J.-C. Thomas, Rational Homotopy Theory, Graduate Texts in Mathematics 205, Springer, New York, 2001.DOI
  14. [14]J. Fine and D. Panov, Circle-invariant fat bundles and symplectic Fano 6-manifolds, J. Lond. Math. Soc. (2) 91 (2015), no. 3, 709–730. doi:10.1112/jlms/jdv011; arXiv:1407.0840.DOI
  15. [15]O. Goertsches, P. Konstantis, and L. Zoller, GKM theory and Hamiltonian non-Kähler actions in dimension 6, Adv. Math. 368 (2020), article 107141, 17 pp. arXiv:1903.11684v2.
  16. [16]O. Goertsches, P. Konstantis, and L. Zoller, Symplectic and Kähler structures on biquotients, J. Symplectic Geom. 18 (2020), no. 3, 791–813. arXiv:1812.09689v2.arxiv.org/abs/1812.09689
  17. [17]K. Grove, Geometry of and via symmetry, in Conformal, Riemannian and Lagrangian Geometry, University Lecture Series 27, American Mathematical Society, Providence, RI, 2002, 31–53.DOI
  18. [18]K. Grove and C. W. Kim, Positively curved manifolds with low fixed point cohomogeneity, J. Differential Geom. 67 (2004), no. 1, 1–33. doi:10.4310/jdg/1099587728.DOI
  19. [19]K. Grove and S. Markvorsen, New extremal problems for the Riemannian recognition program via Alexandrov geometry, J. Amer. Math. Soc. 8 (1995), no. 1, 1–28.
  20. [20]K. Grove and C. Searle, Positively curved manifolds with maximal symmetry-rank, J. Pure Appl. Algebra 91 (1994), no. 1–3, 137–142.arxiv.org/abs/1208.1206
  21. [21]K. Grove and C. Searle, Differential topological restrictions, curvature and symmetry, J. Differential Geom. 47 (1997), no. 3, 530–559.DOI
  22. [22]K. Grove, K. Shankar, and W. Ziller, Symmetries of Eschenburg spaces and the Chern problem, Asian J. Math. 10 (2006), no. 3, 647–662. arXiv:math/0605079.DOI
  23. [23]K. Grove, L. Verdiani, and W. Ziller, An exotic T₁S⁴ with positive curvature, Geom. Funct. Anal. 21 (2011), no. 3, 499–524.
  24. [24]K. Grove and B. Wilking, A knot characterization and 1-connected nonnegatively curved 4-manifolds with circle symmetry, Geom. Topol. 18 (2014), no. 5, 3091–3110.
  25. [25]K. Grove, B. Wilking, and W. Ziller, Positively curved cohomogeneity one manifolds and 3-Sasakian geometry, J. Differential Geom. 78 (2008), no. 1, 33–111. arXiv:math/0511464.
  26. [26]K. Grove and W. Ziller, Curvature and symmetry of Milnor spheres, Ann. of Math. (2) 152 (2000), no. 1, 331–367. arXiv:math/0007198.arxiv.org/abs/math/0007198
  27. [27]S. Guo, E.-X. Fang, and J. Lu, A metric with positive sectional curvature on S² × S³, preprint, arXiv:2608.22133v2, revised August 26, 2026.
  28. [28]S. Guo, E.-X. Fang, and J. Lu, A metric with positive sectional curvature on S³ × S³, preprint, arXiv:2609.11980v1, September 6, 2026.
  29. [29]Y.-H. He, Z. Liu, and S.-T. Yau, Positive sectional curvature on all smooth 7-spheres, preprint, arXiv:2609.29426v1, September 24, 2026.
  30. [30]M. Herrmann, Classification and characterization of rationally elliptic manifolds in low dimensions, Math. Z. 288 (2018), no. 3–4, 1081–1101. doi:10.1007/s00209-017-1927-7; arXiv:1409.8036v2.DOI
  31. [31]W.-Y. Hsiang and B. Kleiner, On the topology of positively curved 4-manifolds with symmetry, J. Differential Geom. 30 (1989), no. 3, 615–621.DOI
  32. [32]D. Husemoller, Fibre Bundles, third edition, Graduate Texts in Mathematics 20, Springer, New York, 1994.
  33. [33]P. E. Jupp, Classification of certain 6-manifolds, Proc. Cambridge Philos. Soc. 73 (1973), no. 2, 293–300. doi:10.1017/S0305004100076854.DOI
  34. [34]L. Kennard, M. Wiemeler, and B. Wilking, Splitting of torus representations and applications in the Grove symmetry program, preprint, arXiv:2106.14723 (2021).arxiv.org/abs/2106.14723
  35. [35]L. Kennard, M. Wiemeler, and B. Wilking, Positive curvature, torus symmetry, and matroids, J. Eur. Math. Soc. (2025), published online first. doi:10.4171/JEMS/1656; arXiv:2212.08152v2.DOI
  36. [36]M. A. Kervaire and J. W. Milnor, Groups of homotopy spheres. I, Ann. of Math. (2) 77 (1963), no. 3, 504–537. doi:10.2307/1970128.DOI
  37. [37]E. Khalili Samani and M. Radeschi, Positive curvature and rational ellipticity in cohomogeneity three, preprint, arXiv:2505.22577v1, May 28, 2025.arxiv.org/abs/2505.22577
  38. [38]A. Krishnan and M. Wiemeler, Positively curved 10-manifolds with T^3-symmetry, Commun. Contemp. Math. 27 (2025), no. 10, 2550018.
  39. [39]Y. Liu, On closed six-manifolds admitting Riemannian metrics with positive sectional curvature and non-abelian symmetry, Acta Math. Sin. (Engl. Ser.) 40 (2024), no. 12, 3003–3026. doi:10.1007/s10114-024-1418-9. For the corrected statements used here, see [40].DOI
  40. [40]Y. Liu, On closed 6-manifolds admitting Riemannian metrics with positive sectional curvature and non-abelian symmetry, corrected manuscript, arXiv:1905.02260v4, revised August 21, 2026. All theorem, remark, and figure numbers attributed to Liu in this paper refer to this version.
  41. [41]J. W. Milnor and J. D. Stasheff, Characteristic Classes, Annals of Mathematics Studies 76, Princeton University Press, Princeton, NJ, 1974.
  42. [42]B. O’Neill, The fundamental equations of a submersion, Michigan Math. J. 13 (1966), no. 4, 459–469. doi:10.1307/mmj/1028999604.DOI
  43. [43]J. Parker, 4-dimensional G-manifolds with 3-dimensional orbits, Pacific J. Math. 125 (1986), no. 1, 187–204.
  44. [44]P. Petersen, Riemannian Geometry, third edition, Graduate Texts in Mathematics 171, Springer, Cham, 2016. doi:10.1007/978-3-319-26654-1.
  45. [45]A. Petrunin, Applications of quasigeodesics and gradient curves, in Comparison Geometry, Mathematical Sciences Research Institute Publications 30, Cambridge University Press, Cambridge, 1997, 203–219.DOI
  46. [46]X. Rong, Positively curved manifolds with almost maximal symmetry rank, Geom. Dedicata 95 (2002), no. 1, 157–182.DOI
  47. [47]F. Simas, Nonnegatively curved five-manifolds with non-abelian symmetry, Geom. Dedicata 181 (2016), 61–82.DOI
  48. [48]S. Smale, Diffeomorphisms of the 2-sphere, Proc. Amer. Math. Soc. 10 (1959), 621–626. doi:10.1090/S0002-9939-1959-0112149-8.DOI
  49. [49]L. Verdiani, Cohomogeneity one manifolds of even dimension with strictly positive sectional curvature, J. Differential Geom. 68 (2004), no. 1, 31–72. doi:10.4310/jdg/1102536709.DOI
  50. [50]C. T. C. Wall, Classification problems in differential topology. V. On certain 6-manifolds, Invent. Math. 1 (1966), 355–374; corrigendum, 2 (1966), 306.
  51. [51]N. R. Wallach, Compact homogeneous Riemannian manifolds with strictly positive curvature, Ann. of Math. (2) 96 (1972), no. 2, 277–295.DOI
  52. [52]B. Wilking, Torus actions on manifolds of positive sectional curvature, Acta Math. 191 (2003), no. 2, 259–297. doi:10.1007/BF02392966.DOI
  53. [53]B. Wilking, Positively curved manifolds with symmetry, Ann. of Math. (2) 163 (2006), no. 2, 607–668.DOI
  54. [54]B. Wilking, Nonnegatively and positively curved manifolds, Surv. Differ. Geom. 11 (2007), 25–62.arxiv.org/abs/0707.3091
  55. [55]W. Ziller, Examples of Riemannian manifolds with non-negative sectional curvature, Surv. Differ. Geom. 11 (2007), 63–102.arxiv.org/abs/math/0701389
  56. [56]W. Ziller, On M. Müter’s Ph.D. thesis on Cheeger deformations, notes by C. He, arXiv:0909.0161 (2009).arxiv.org/abs/0909.0161
  57. [57]A. V. Žubr, Closed simply connected six-dimensional manifolds: proofs of classification theorems, Algebra i Analiz 12 (2000), no. 4, 126–230; English translation, St. Petersburg Math. J. 12 (2001), no. 4, 605–680.

Paper details

vlnc:W_hghxjuuv
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

ChatGPT-6-astra

View statement
Looking for related papers…
Contents